Performance of comorbidity scores to control for confounding in epidemiologic studies using claims data

Multimorbidity prevalence and health outcome prediction: assessing the impact of lookback periods, disease count, and definition criteria in health administrative data at the population-based level

BACKGROUND

Health administrative databases play a crucial role in population-level multimorbidity surveillance. Determining the appropriate retrospective or lookback period (LP) for observing prevalent and newly diagnosed diseases in administrative data presents challenge in estimating multimorbidity prevalence and predicting health outcome. The aim of this population-based study was to assess the impact of LP on multimorbidity prevalence and health outcomes prediction across three multimorbidity definitions, three lists of diseases used for multimorbidity assessment, and six health outcomes.

METHODS

We conducted a population-based study including all individuals ages > 65 years on April 1st, 2019, in Québec, Canada. We considered three lists of diseases labeled according to the number of chronic conditions it considered: (1) L60 included 60 chronic conditions from the International Classification of Diseases (ICD); (2) L20 included a core of 20 chronic conditions; and (3) L31 included 31 chronic conditions from the Charlson and Elixhauser indices. For each list, we: (1) measured multimorbidity prevalence for three multimorbidity definitions (at least two [MM2+], three [MM3+] or four (MM4+) chronic conditions); and (2) evaluated capacity (c-statistic) to predict 1-year outcomes (mortality, hospitalisation, polypharmacy, and general practitioner, specialist, or emergency department visits) using LPs ranging from 1 to 20 years.

RESULTS

Increase in multimorbidity prevalence decelerated after 5-10 years (e.g., MM2+, L31: LP = 1y: 14%, LP = 10y: 58%, LP = 20y: 69%). Within the 5-10 years LP range, predictive performance was better for L20 than L60 (e.g., LP = 7y, mortality, MM3+: L20 [0.798;95%CI:0.797-0.800] vs. L60 [0.779; 95%CI:0.777-0.781]) and typically better for MM3 + and MM4 + definitions (e.g., LP = 7y, mortality, L60: MM4+ [0.788;95%CI:0.786-0.790] vs. MM2+ [0.768;95%CI:0.766-0.770]).

CONCLUSIONS

In our databases, ten years of data was required for stable estimation of multimorbidity prevalence. Within that range, the L20 and multimorbidity definitions MM3 + or MM4 + reached maximal predictive performance.

Cardiovascular and mortality benefits of sodium-glucose cotransporter-2 inhibitors and glucagon-like peptide 1 receptor agonists as third-step glucose-lowering medicine in patients with type 2 diabetes: a retrospective cohort analysis

INTRODUCTION

Studies have found that sodium-glucose cotransporter 2 inhibitors (SGLT2) and glucagon-like peptide 1 receptor agonists (GLP1) have cardiovascular benefits for patients with type 2 diabetes (DM2) and atherosclerotic cardiovascular disease (ASCVD), chronic kidney disease (CKD), or heart failure (HF). The literature does not provide evidence specifically for patients with these conditions who are adding one of these medicines to two glucose-lowering medications (ie, as "third-step" therapy). We explored the effects of different third-step medications on cardiovascular outcomes in patients with diabetes and these comorbid conditions. Specifically, we compared third-step SGLT2 or GLP1 to third-step dipeptidyl peptidase-4 inhibitors (DPP4), insulin, or thiazolidinediones (TZD).

RESEARCH DESIGN AND METHODS

We assembled a retrospective cohort of adults at five Kaiser Permanente sites with DM2 and ASCVD, CKD, or HF, initiating third-step treatment between 2016 and 2020. Propensity score weighted Poisson models were used to calculate adjusted rate ratios (ARRs) for all-cause mortality, incident major adverse cardiovascular event (MACE), and incident HF hospitalization in patients initiating SGLT2 or GLP1 compared with DPP4, insulin, or TZD.

RESULTS

We identified 27 542 patients initiating third-step treatment with one or more of these conditions (19 958 with ASCVD, 14 577 with CKD, and 3919 with HF). ARRs for GLP1 and SGLT2 versus DPP4, insulin, and TZD in the patient subgroups ranged between 0.22 and 0.55 for all-cause mortality, 0.38 and 0.81 for MACE, and 0.46 and 1.05 for HF hospitalization. Many ARRs were statistically significant, and all significant ARRs showed a benefit (ARR <1) for GLP1 or SGLT2 when compared with DPP4, insulin, or TZD.

CONCLUSIONS

Third-step SGLT2 and GLP1 are generally associated with a benefit for these outcomes in these patient groups when compared with third-step DPP4, insulin, or TZD. Our results add to evidence of a cardiovascular benefit of SGLT2 and GLP1 and could inform clinical guidelines for choosing third-step diabetes treatment.

Association between surgery and rate of incident dementia in older adults: A population-based retrospective cohort study

BACKGROUND

The risk of incident dementia after surgery in older adults is unclear. The study objective was to examine the rate of incident dementia among older adults after elective surgery compared with a matched nonsurgical control group.

METHODS

We conducted a population-based, propensity-matched retrospective cohort study using data from linked administrative databases in Ontario, Canada. All community-dwelling individuals aged 66 years and older who underwent one of five major elective surgeries between April 1, 2007 and March 31, 2011 were included. Each surgical patient was matched 1:1 on surgical specialty of the surgeon at consultation, age, sex, fiscal year of entry, and propensity score with a patient who attended an outpatient visit with a surgeon of the same surgical specialty but did not undergo surgery. Patients were followed for up to 5 years after cohort entry for the occurrence of a new dementia diagnosis, defined from administrative data. Cause-specific hazard models were used to estimate the hazard ratio (HR) and 95% confidence interval (CI) for the association between surgery and the hazard of incident dementia. Subgroup and sensitivity analyses were performed.

RESULTS

A total of 27,878 individuals (13,939 matched pairs) were included in the analysis. A total of 640 (4.6%) individuals in the surgical group and 965 (6.9%) individuals in the control group developed dementia over the 5-year follow-up period. Individuals who underwent surgery had a reduced rate of incident dementia compared with their matched nonsurgical controls (HR 0.88; 95% CI 0.80-0.97; p = 0.01). This association was persistent in most subgroups and after sensitivity analyses.

CONCLUSIONS

Elective surgery did not increase the rate of incident dementia when compared with matched nonsurgical controls. This could be an important consideration for patients and surgeons when elective surgery is considered.

The Postoperative Airway Compromise Score-First Steps to Developing a Postoperative Tool for the Assessment of Upper Airway-Related Complications Following Anterior Cervical Spine Surgery

BACKGROUND

Acute upper airway compromise is a rare but catastrophic complication after anterior cervical discectomy and fusion. This study aims to develop a score to identify patients at risk of acute postoperative airway compromise (PAC).

METHODS

Potential risk factors for acute PAC were selected by a modified Delphi process. Ten patients with acute PAC were identified of 1466 patients who underwent elective anterior cervical discectomy and fusion between July 2014 and May 2019. A comparison group was created by a randomized selection process (non-PAC group). Factors associated with PAC and a P value of < 0.10 were entered into a logistic regression model and coefficients contributed to each risk factor's overall score. Calibration of the model was evaluated using the Hosmer-Lemeshow goodness-of-fit test. Quantitative discrimination was calculated, and the final model was internally validated with bootstrap sampling.

RESULTS

We identified 18 potential risk factors from our Delphi process, of which 6 factors demonstrated a significant association with airway compromise: age >65 years, current smoking status, American Society of Anesthesiologists class >2, history of a bleeding disorder, surgery of upper subaxial cervical spine (above C4), and duration of surgery >179 minutes. The final prediction model included 5 predictors with very strong performance characteristics. These 5 factors formed the PAC score, with a range from 0 to 100. A score of 20 yielded the greatest balance of sensitivity (80%) and specificity (88%).

CONCLUSIONS

The acute PAC score demonstrates strong performance characteristics. The PAC score might help identify patients at risk of upper airway compromise caused by surgical site abnormalities.

The medication-based Rx-Risk Comorbidity Index and risk of hip fracture - a nationwide NOREPOS cohort study

BACKGROUND

Few previous studies have assessed overall morbidity at the individual level with respect to future risk of hip fracture. The aim of this register-based cohort study was to examine the association between morbidity measured by the medication-based Rx-Risk Comorbidity Index (Rx-Risk) and the risk of first hip fracture.

METHODS

Individual-level data on medications dispensed from pharmacies (2005-2016) was retrieved from the Norwegian Prescription Database and used to calculate Rx-Risk for each calendar year. Information on first hip fractures (2006-2017) was obtained from a nationwide hip fracture database. Individuals ≥ 51 years who filled at least one prescription during the study period comprised the population at risk. Using Rx-Risk as a time-varying exposure variable, relative risk estimates were obtained by a negative binomial model.

RESULTS

During 2006-2017, 94,104 individuals sustained a first hip fracture. A higher Rx-Risk was associated with increased risk of hip fracture within all categories of age and sex. Women with the highest Rx-Risk (> 25) had a relative risk of 6.1 (95% confidence interval (CI): 5.4, 6.8) compared to women with Rx-Risk ≤ 0, whereas the corresponding relative risk in women with Rx-Risk 1-5 was 1.4 (95% CI: 1.3, 1.4). Similar results were found in men. Women > 80 years with Rx-Risk 21-25 had the highest incidence rate (514 (95% CI: 462, 566) per 10, 000 person years). The relative increase in hip fracture risk with higher Rx-Risk was most pronounced in the youngest patients aged 51-65 years.

CONCLUSIONS

Rx-Risk is a strong predictor of hip fracture in the general outpatient population and may be useful to identify individuals at risk in a clinical setting and in future studies.

Association of polypharmacy scores with the long-term survival of patients with intact aortoiliac aneurysms and indication for repair

OBJECTIVE/BACKGROUND

Our study analyzed the relationship between two polypharmacy scores (addition of chronic prescribed drugs [ACPDs] and Rx-Risk Comorbidity Index) and survival in patients with an intact abdominal aortic and/or common iliac aneurysm (AAA).

METHODS

Consecutive retrospective, single-center cohort of patients attended for an intact AAA with indication for repair from 2008 to 2021. Demographic data, Charlson Comorbidity Index, AAA treatment, ACPD, and Rx-Risk polypharmacy scores were recorded at baseline. Main outcomes were the 5-year and long-term survival rates. The statistical analysis included Cox regression, area under the curve, and continuous net reclassification index.

RESULTS

A total of 424 patients with AAA were evaluated (median age: 76 years; 92.2% male, median Charlson index 2), of whom 314 (74.1%) underwent intervention (80% endovascular and 20% open) and 110 (25.9%) did not. During follow-up (mean 4.6 years), 245 patients (57.8%) died, with 1-month, 1-year, and 5-year survival rates of 98.1%, 86.3%, and 52.7%, respectively. ACPD and Rx-Risk indices (median [interquartile range]: 6 [4-9] and 3 [0-5], respectively) were significantly and linearly associated (P < .001) with survival, with the best cutoff points at 5 and 0, respectively. An ACPD >5 (patients with >5 chronically prescribed drugs at baseline) and an Rx-Risk >0 were associated with a 45.2% (P = .038) and 102% (P = .002) increase in 5-year mortality, respectively, after adjustment for age, sex, Charlson index, and type of AAA treatment. Both polypharmacy indices improved significantly the discriminative power of the Charlson Comorbidity Index in predicting survival.

CONCLUSIONS

Both ACPD and Rx-Risk polypharmacy scores are independently related to survival among patients with an intact AAA and indication for repair. Their behavior is similar, so the simple ACPD >5 appears to be sufficient to identify patients with lower survival rates.

Gabapentinoids and Risk for Severe Exacerbation in Chronic Obstructive Pulmonary Disease : A Population-Based Cohort Study

BACKGROUND

North American and European health agencies recently warned of severe breathing problems associated with gabapentinoids, including in patients with chronic obstructive pulmonary disease (COPD), although supporting evidence is limited.

OBJECTIVE

To assess whether gabapentinoid use is associated with severe exacerbation in patients with COPD.

DESIGN

Time-conditional propensity score-matched, new-user cohort study.

SETTING

Health insurance databases from the Régie de l'assurance maladie du Québec in Canada.

PATIENTS

Within a base cohort of patients with COPD between 1994 and 2015, patients initiating gabapentinoid therapy with an indication (epilepsy, neuropathic pain, or other chronic pain) were matched 1:1 with nonusers on COPD duration, indication for gabapentinoids, age, sex, calendar year, and time-conditional propensity score.

MEASUREMENTS

The primary outcome was severe COPD exacerbation requiring hospitalization. Hazard ratios (HRs) associated with gabapentinoid use were estimated in subcohorts according to gabapentinoid indication and in the overall cohort.

RESULTS

The cohort included 356 gabapentinoid users with epilepsy, 9411 with neuropathic pain, and 3737 with other chronic pain, matched 1:1 to nonusers. Compared with nonuse, gabapentinoid use was associated with increased risk for severe COPD exacerbation across the indications of epilepsy (HR, 1.58 [95% CI, 1.08 to 2.30]), neuropathic pain (HR, 1.35 [CI, 1.24 to 1.48]), and other chronic pain (HR, 1.49 [CI, 1.27 to 1.73]) and overall (HR, 1.39 [CI, 1.29 to 1.50]).

LIMITATION

Residual confounding, including from lack of smoking information.

CONCLUSION

In patients with COPD, gabapentinoid use was associated with increased risk for severe exacerbation. This study supports the warnings from regulatory agencies and highlights the importance of considering this potential risk when prescribing gabapentin and pregabalin to patients with COPD.

PRIMARY FUNDING SOURCE

Canadian Institutes of Health Research and Canadian Lung Association.

The Patient Generated Index (PGI) as an early-warning system for predicting brain health challenges: a prospective cohort study for people living with Human Immunodeficiency Virus (HIV)

PURPOSE

In research people are often asked to fill out questionnaires about their health and functioning and some of the questions refer to serious health concerns. Typically, these concerns are not identified until the statistician analyses the data. An alternative is to use an individualized measure, the Patient Generated Index (PGI) where people are asked to self-nominate areas of concern which can then be dealt with in real-time. This study estimates the extent to which self-nominated areas of concern related to mood, anxiety and cognition predict the presence or occurrence of brain health outcomes such as depression, anxiety, psychological distress, or cognitive impairment among people aging with HIV at study entry and for successive assessments over 27 months.

METHODS

The data comes from participants enrolled in the Positive Brain Health Now (+ BHN) cohort (n = 856). We analyzed the self-nominated areas that participants wrote on the PGI and classified them into seven sentiment groups according to the type of sentiment expressed: emotional, interpersonal, anxiety, depressogenic, somatic, cognitive and positive sentiments. Tokenization was used to convert qualitative data into quantifiable tokens. A longitudinal design was used to link these sentiment groups to the presence or emergence of brain health outcomes as assessed using standardized measures of these constructs: the Hospital Anxiety and Depression Scale (HADS), the Mental Health Index (MHI) of the RAND-36, the Communicating Cognitive Concerns Questionnaire (C3Q) and the Brief Cognitive Ability Measure (B-CAM). Logistic regressions were used to estimate the goodness of fit of each model using the c-statistic.

RESULTS

Emotional sentiments predicted all of the brain health outcomes at all visits with adjusted odds ratios (OR) ranging from 1.61 to 2.00 and c-statistics > 0.73 (good to excellent prediction). Nominating an anxiety sentiment was specific to predicting anxiety and psychological distress (OR 1.65 & 1.52); nominating a cognitive concern was specific to predicting self-reported cognitive ability (OR 4.78). Positive sentiments were predictive of good cognitive function (OR 0.36) and protective of depressive symptoms (OR 0.55).

CONCLUSIONS

This study indicates the value of using this semi-qualitative approach as an early-warning system in predicting brain health outcomes.

Variability in changes in physician outpatient antibiotic prescribing from 2019 to 2021 during the COVID-19 pandemic in Ontario, Canada

Objective

To evaluate inter-physician variability and predictors of changes in antibiotic prescribing before (2019) and during (2020/2021) the coronavirus disease 2019 (COVID-19) pandemic.

Methods

We conducted a retrospective cohort analysis of physicians in Ontario, Canada prescribing oral antibiotics in the outpatient setting between January 1, 2019 and December 31, 2021 using the IQVIA Xponent data set. The primary outcome was the change in the number of antibiotic prescriptions between the prepandemic and pandemic period. Secondary outcomes were changes in the selection of broad-spectrum agents and long-duration (>7 d) antibiotic use. We used multivariable linear regression models to evaluate predictors of change.

Results

There were 17,288 physicians included in the study with substantial inter-physician variability in changes in antibiotic prescribing (median change of -43.5 antibiotics per physician, interquartile range -136.5 to -5.0). In the multivariable model, later career stage (adjusted mean difference [aMD] -45.3, 95% confidence interval [CI] -52.9 to -37.8, p < .001), family medicine (aMD -46.0, 95% CI -62.5 to -29.4, p < .001), male patient sex (aMD -52.4, 95% CI -71.1 to -33.7, p < .001), low patient comorbidity (aMD -42.5, 95% CI -50.3 to -34.8, p < .001), and high prescribing to new patients (aMD -216.5, 95% CI -223.5 to -209.5, p < .001) were associated with decreases in antibiotic initiation. Family medicine and high prescribing to new patients were associated with a decrease in selection of broad-spectrum agents and prolonged antibiotic use.

Conclusions

Antibiotic prescribing changed throughout the COVID-19 pandemic with overall decreases in antibiotic initiation, broad-spectrum agents, and prolonged antibiotic courses with inter-physician variability. These findings present opportunities for community antibiotic stewardship interventions.

Executive function and episodic memory composite scores in older adults: relations with sex, mood, and subjective sleep quality

Executive function and episodic memory processes are particularly vulnerable to aging. We sought to learn the degree to which sex, mood, and subjective sleep quality might be related to executive function and episodic memory composite scores in community-dwelling older adults. We replicated Glisky and colleagues' two-factor (i.e., executive function [N=263] versus episodic memory [N=151]) structure, and found that it did not significantly differ between males and females. Moderation analyses revealed no interactions between sex, mood, and sleep in predicting either composite score. However, females significantly outperformed males on the episodic memory composite, and on all the individual tests comprising it. Ours is the first study to look at sex differences in this battery's factor structure and its potential relations with mood and sleep. Future longitudinal studies in both healthy and clinical populations will help us further probe the possible influence of these variables on executive function and episodic memory in aging.

Association between pre-diagnosis geriatric syndromes and overall survival in older adults with cancer (the INCAPAC study)

INTRODUCTION

Several population-based studies have reported disparities in overall survival (OS) among older patients with cancer. However, geriatric syndromes, known to be associated with OS in the geriatric population, were rarely studied. Thus, our aim was to identify the determinants of OS among French older adults with cancer, including geriatric syndromes before cancer diagnosis.

MATERIALS AND METHODS

Using cancer registries, we identified older subjects (≥65 years) with cancer in three French prospective cohort studies on aging from the Gironde department. Survival time was calculated from the date of diagnosis to the date of all-cause death or to the date of last follow-up, whichever came first. Demographic and socioeconomic characteristics, smoking status, self-rated health, cancer-related factors (stage at diagnosis, treatment), as well as geriatric syndromes (polypharmacy, activity limitation, depressive symptomatology, and cognitive impairment or dementia) were studied. Analyses were performed using Cox proportional hazard models for the whole population, then by age group (65-84 and 85+).

RESULTS

Among the 607 subjects included in the study, the median age at cancer diagnosis was 84 years. Smoking habits, activity limitations, cognitive impairment or dementia, advanced cancer stage and absence of treatment were significantly associated with lower OS in the analysis including the whole population. Women presented higher OS. Factors associated with OS differed by age group. Polypharmacy was inversely associated with OS in older adults aged 65-84 and 85 + .

DISCUSSION

Our findings support that geriatric assessment is needed to identify patients at higher risk of death and that an evaluation of activity limitation in older adults is essential. Improving early detection could enable interventions to address factors (activity limitations, cognitive impairment) associated with OS, potentially reducing disparities and lead to earlier palliative care.

Hydroxychloroquine lowers Alzheimer's disease and related dementias risk and rescues molecular phenotypes related to Alzheimer's disease

We recently nominated cytokine signaling through the Janus-kinase-signal transducer and activator of transcription (JAK/STAT) pathway as a potential AD drug target. As hydroxychloroquine (HCQ) has recently been shown to inactivate STAT3, we hypothesized that it may impact AD pathogenesis and risk. Among 109,124 rheumatoid arthritis patients from routine clinical care, HCQ initiation was associated with a lower risk of incident AD compared to methotrexate initiation across 4 alternative analyses schemes addressing specific types of biases including informative censoring, reverse causality, and outcome misclassification (hazard ratio [95% confidence interval] of 0.92 [0.83-1.00], 0.87 [0.81-0.93], 0.84 [0.76-0.93], and 0.87 [0.75-1.01]). We additionally show that HCQ exerts dose-dependent effects on late long-term potentiation (LTP) and rescues impaired hippocampal synaptic plasticity prior to significant accumulation of amyloid plaques and neurodegeneration in APP/PS1 mice. Additionally, HCQ treatment enhances microglial clearance of Aβ_1-42, lowers neuroinflammation, and reduces tau phosphorylation in cell culture-based phenotypic assays. Finally, we show that HCQ inactivates STAT3 in microglia, neurons, and astrocytes suggesting a plausible mechanism associated with its observed effects on AD pathogenesis. HCQ, a relatively safe and inexpensive drug in current use may be a promising disease-modifying AD treatment. This hypothesis merits testing through adequately powered clinical trials in at-risk individuals during preclinical stages of disease progression.

High-dimensional propensity scores for empirical covariate selection in secondary database studies: Planning, implementation, and reporting

Real-world evidence used for regulatory, payer, and clinical decision-making requires principled epidemiology in design and analysis, applying methods to minimize confounding given the lack of randomization. One technique to deal with potential confounding is propensity score (PS) analysis, which allows for the adjustment for measured preexposure covariates. Since its first publication in 2009, the high-dimensional propensity score (hdPS) method has emerged as an approach that extends traditional PS covariate selection to include large numbers of covariates that may reduce confounding bias in the analysis of healthcare databases. hdPS is an automated, data-driven analytic approach for covariate selection that empirically identifies preexposure variables and proxies to include in the PS model. This article provides an overview of the hdPS approach and recommendations on the planning, implementation, and reporting of hdPS used for causal treatment-effect estimations in longitudinal healthcare databases. We supply a checklist with key considerations as a supportive decision tool to aid investigators in the implementation and transparent reporting of hdPS techniques, and to aid decision-makers unfamiliar with hdPS in the understanding and interpretation of studies employing this approach. This article is endorsed by the International Society for Pharmacoepidemiology.

Proton pump inhibitors and the risk of inflammatory bowel disease: population-based cohort study

OBJECTIVE

To determine whether the use of proton pump inhibitors (PPIs) compared with the use of histamine-2 receptor antagonists (H2RAs) is associated with an increased risk of inflammatory bowel disease (IBD).

DESIGN

Population-based cohort study designed to address the impact of protopathic bias.

SETTING

General practices contributing data to the UK Clinical Practice Research Datalink GOLD.

PARTICIPANTS

1 498 416 initiators of PPIs and 322 474 initiators of H2RAs from 1 January 1990 to 31 December 2018, with follow-up until 31 December 2019. Patients were analysed according to the timing of the IBD diagnosis after treatment initiation (early vs late).

MAIN OUTCOME MEASURES

Standardised morbidity ratio weighted Cox proportional hazards models were used to estimate marginal HRs and 95% CIs. In the early-event analysis, IBD diagnoses were assessed within the first 2 years of treatment initiation, an analysis subject to potential protopathic bias. In the late-event analysis, all exposures were lagged by 2 years to account for latency and minimise protopathic bias.

RESULTS

In the early-event analysis, the use of PPIs was associated with an increased risk of IBD within the first 2 years of treatment initiation, compared with H2RAs (HR 1.39, 95% CI 1.14 to 1.69). In contrast, the use of PPIs was not associated with an increased risk of IBD in the late-event analysis (HR 1.05, 95% CI 0.90 to 1.22). The results remained consistent in several sensitivity analyses.

CONCLUSIONS

Compared with H2RAs, PPIs were not associated with an increased risk of IBD, after accounting for protopathic bias.

Medication Prescribed Within One Year Preceding Fall-Related Injuries in Ontario Older Adults

Background

Serious injuries secondary to falls are becoming more prevalent due to the worldwide ageing of societies. Several medication classes have been associated with falls and fall-related injuries. The purpose of this study was to describe medication classes and the number of medication classes prescribed to older adults prior to the fall-related injury.

Methods

This population-based descriptive study used secondary administrative health-care data in Ontario, Canada for 2010-2014. Descriptive statistics were reported for Anatomic Therapeutic Chemical 4^th level medication classes. Frequency of medications prescribed to older adults was calculated on different sex, age groups, types of medications, and injures.

Results

Over five years (2010-2014), 288,251 older adults (63.2% females) were admitted to an emergency department for a fall-related injury (40.0% fractures, 12.1% brain injury). In the year before the injury, 48.5% were prescribed statins, 27.2% antidepressants, 25.0% opioids, and 16.6% anxiolytics. Females were prescribed more diuretics, antidepressants, and anxiolytics than males; and people aged 85 years and older had a higher percentage of diuretics, antidepressants, and antipsychotics. There were 36.4% of older adults prescribed 5-9 different medication classes and 41.2% were prescribed 10 or more medication classes.

Discussion

Older adults experiencing fall-related injuries were prescribed more opioids, benzodiazepines, and antidepressants than previously reported for the general population of older adults in Ontario. Higher percentage of females and more 85+ older adults were prescribed with psychotropic drugs, and they were also found to be at higher risk of fall-related injuries. Further associations between medications and fall-related injuries need to be explored in well-defined cohort studies.

Development and Validation of the Summary Elixhauser Comorbidity Score for Use With ICD-10-CM-Coded Data Among Older Adults

BACKGROUND

Older adults have many comorbidities contributing to mortality.

OBJECTIVE

To develop a summary Elixhauser (S-Elixhauser) comorbidity score to predict 30-day, in-hospital, and 1-year mortality in older adults using the 38 comorbidities operationalized by the Agency for Healthcare Research and Quality (AHRQ).

DESIGN

Retrospective cohort study.

SETTING

Medicare beneficiaries from 2017 to 2019.

PATIENTS

Persons hospitalized in 2018 (n = 899 844) and 3 disease-specific hospitalized cohorts.

MEASUREMENTS

Weights were derived for 38 comorbidities to predict 30-day, in-hospital, and 1-year mortality. The S-Elixhauser score was internally validated and calibrated. Individual Elixhauser comorbidity indicators (38 comorbidities), the modified application of the AHRQ-derived Elixhauser summary score, the Charlson comorbidity indicators (17 comorbidities), and the Charlson summary score were externally validated. The c-statistic was used to evaluate discrimination of a comorbidity score model.

RESULTS

The S-Elixhauser score was well calibrated and internally validated, with a c-statistic of 0.705 (95% CI, 0.703 to 0.707) in predicting 30-day mortality, 0.654 (CI, 0.651 to 0.657) for in-hospital mortality, and 0.743 (CI, 0.741 to 0.744) for 1-year mortality. In external validation of other comorbidity indices for 30-day mortality, the c-statistic was 0.711 (CI, 0.709 to 0.713) for the individual Elixhauser comorbidity indicators, 0.688 (CI, 0.686 to 0.690) for the AHRQ Elixhauser score, 0.696 (CI, 0.694 to 0.698) for the Charlson comorbidity indicators, and 0.690 (CI, 0.688 to 0.693) for the Charlson summary score. In 3 disease-specific populations, the discrimination of the S-Elixhauser score in predicting 30-day mortality ranged from 0.657 to 0.732.

LIMITATION

Validation of the S-Elixhauser comorbidity score and head-to-head comparison with other comorbidity scores in an external population are needed to evaluate comparative performance.

CONCLUSION

The S-Elixhauser comorbidity score is well calibrated and internally validated but its advantage over the AHRQ Elixhauser and Charlson summary scores is unclear.

PRIMARY FUNDING SOURCE

National Institute on Aging.

Utility of automated data-adaptive propensity score method for confounding by indication in comparative effectiveness study in real world Medicare and registry data

Background

Confounding by indication is a serious threat to comparative studies using real world data. We assessed the utility of automated data-adaptive analytic approach for confounding adjustment when both claims and clinical registry data are available.

Methods

We used a comparative study example of carotid artery stenting (CAS) vs. carotid endarterectomy (CEA) in 2005–2008 when CAS was only indicated for patients with high surgical risk. We included Medicare beneficiaries linked to the Society for Vascular Surgery’s Vascular Registry >65 years old undergoing CAS/CEA. We compared hazard ratios (HRs) for death while adjusting for confounding by combining various 1) Propensity score (PS) modeling strategies (investigator-specified [IS-PS] vs. automated data-adaptive [ada-PS]); 2) data sources (claims-only, registry-only and claims-plus-registry); and 3) PS adjustment approaches (matching vs. quintiles-adjustment with/without trimming). An HR of 1.0 was used as a benchmark effect estimate based on CREST trial.

Results

The cohort included 1,999 CAS and 3,255 CEA patients (mean age 76). CAS patients were more likely symptomatic and at high surgical risk, and experienced higher mortality (crude HR = 1.82 for CAS vs. CEA). HRs from PS-quintile adjustment without trimming were 1.48 and 1.52 for claims-only IS-PS and ada-PS, 1.51 and 1.42 for registry-only IS-PS and ada-PS, and 1.34 and 1.23 for claims-plus-registry IS-PS and ada-PS, respectively. Estimates from other PS adjustment approaches showed similar patterns.

Conclusions

In a comparative effectiveness study of CAS vs. CEA with strong confounding by indication, ada-PS performed better than IS-PS in general, but both claims and registry data were needed to adequately control for bias.

A National Catalog of Mapped Short-Form Six-Dimension Utility Scores for Chronic Conditions in the United States From 2010 to 2015

OBJECTIVES

This study examined health preference utility weights and utility decrements associated with different types of chronic conditions in the United States.

METHODS

We used the 2010-2015 Medical Expenditure Panel Survey data for persons aged ≥ 18 years with 12-Item Short-Form Survey Physical and Mental Component Summary scores. 12-Item Short-Form Survey scores were converted to Short-Form Six-Dimension (SF-6D) preference scores to measure utilities of different chronic diseases. We used the Clinical Classification Code to identify 30 chronic diseases from 12 categories, such as cardiovascular diseases, cerebrovascular diseases, hypertension, hyperlipidemia, obesity, cancers, musculoskeletal diseases, endocrine or metabolic diseases, oral diseases, respiratory diseases, and mental disorders. A generalized linear model was used to quantify the utility decrements for 30 chronic diseases, controlling for demographic characteristics.

RESULTS

We identified 132 737 adults (mean age 47.2 years, 52.2% female, 80% white); 73% had at least one identified chronic disease, and the mean SF-6D was 0.786. Among 30 chronic diseases, the unadjusted mean SF-6D scores of patients with cognitive disorder (0.607) were the lowest, followed by congestive heart failure (0.629), rheumatoid arthritis (0.654), and lung cancer (0.662). After controlling for demographic variables (ie, age, sex) and comorbidities, cognitive disorders (-0.116), mood disorders (-0.099), rheumatoid arthritis (-0.090), liver cancer (-0.078), and stroke (-0.063) showed the highest decrements in the SF-6D scores (P < .05).

CONCLUSIONS

This study provides a nationally representative catalog of utility weights for major chronic diseases in the US general population. The utility decrements will enable researchers to calculate the health utilities of patients with multiple comorbid diseases.

Impact of the Canadian CT head rule supplemented by the original published minimum inclusion criteria to assist emergency department clinicians' assessment of patients presenting post fall from residential aged care: a retrospective audit

Background

A large number of CT brain (CTB) scans are ordered in the ED for older patients with a confirmed or possible head strike but no ongoing symptoms of a head injury. This study aimed to evaluate the effect of the Canadian CT head rule supplemented by the original published minimum inclusion criteria to assist clinician assessment of the need for CTB following minimal trauma fall in patients presenting from residential aged care facilities to a major metropolitan emergency department (ED).

Methods

This study was conducted as a pre- and post-intervention retrospective audit. The intervention involved implementation of a decision support tool to help clinicians assess patients presenting to the ED following a fall. The tool integrated the Canadian CT Head Rule (CCHR) in conjunction with a simplified set of inclusion criteria to help clinicians define a minimum threshold for a “minor head injury”. Outcome data pertaining to CT brain ordering practices and results were compared over symmetrical 3-month time periods pre- and post-intervention in 2 consecutive years.

Results

The study included 233 patients in the pre-intervention arm and 241 in the post-intervention arm. Baseline demographics and clinical characteristics were similar in both groups. There was a 20% reduction in the total number of CTB scans ordered following tool implementation, with 134 (57.0%) scans in the pre-intervention group and 90 (37.3%) in the post-intervention group (p <  0.01). The diagnostic yield in the pre- and post-intervention groups was 3.7 and 5.6% respectively (p = 0.52). No variation was observed in medical management between groups, and no patients in either group underwent neurosurgical intervention.

Conclusions

Use of the CCHR supplemented by the original published minimum inclusion criteria appeared to safely reduce the number of CTB scans performed in residential aged care facility residents presenting to an ED after a fall, with no associated adverse outcomes. A larger study across multiple centres is required to determine widespread efficacy and safety of this tool.

Individual dynamic prediction of clinical endpoint from large dimensional longitudinal biomarker history: a landmark approach

Background

The individual data collected throughout patient follow-up constitute crucial information for assessing the risk of a clinical event, and eventually for adapting a therapeutic strategy. Joint models and landmark models have been proposed to compute individual dynamic predictions from repeated measures to one or two markers. However, they hardly extend to the case where the patient history includes much more repeated markers. Our objective was thus to propose a solution for the dynamic prediction of a health event that may exploit repeated measures of a possibly large number of markers.

Methods

We combined a landmark approach extended to endogenous markers history with machine learning methods adapted to survival data. Each marker trajectory is modeled using the information collected up to the landmark time, and summary variables that best capture the individual trajectories are derived. These summaries and additional covariates are then included in different prediction methods adapted to survival data, namely regularized regressions and random survival forests, to predict the event from the landmark time. We also show how predictive tools can be combined into a superlearner. The performances are evaluated by cross-validation using estimators of Brier Score and the area under the Receiver Operating Characteristic curve adapted to censored data.

Results

We demonstrate in a simulation study the benefits of machine learning survival methods over standard survival models, especially in the case of numerous and/or nonlinear relationships between the predictors and the event. We then applied the methodology in two prediction contexts: a clinical context with the prediction of death in primary biliary cholangitis, and a public health context with age-specific prediction of death in the general elderly population.

Conclusions

Our methodology, implemented in R, enables the prediction of an event using the entire longitudinal patient history, even when the number of repeated markers is large. Although introduced with mixed models for the repeated markers and methods for a single right censored time-to-event, the technique can be used with any other appropriate modeling technique for the markers and can be easily extended to competing risks setting.

Supplementary Information

The online version contains supplementary material available at (10.1186/s12874-022-01660-3).

Evaluation of risk adjustment performance of diagnosis-based and medication-based comorbidity indices in patients with chronic obstructive pulmonary disease

Objectives

This study assessed risk adjustment performance of six comorbidity indices in two categories of comorbidity measures: diagnosis-based comorbidity indices and medication-based ones in patients with chronic obstructive pulmonary disease (COPD).

Methods

This was a population–based retrospective cohort study. Data used in this study were sourced from the Taiwan National Health Insurance Research Database. The study population comprised all patients who were hospitalized due to COPD for the first time in the target year of 2012. Each qualified patient was individually followed for one year starting from the index date to assess two outcomes of interest, medical expenditures within one year after discharge and in-hospital mortality of patients. To assess how well the added comorbidity measures would improve the fitted model, we calculated the log-likelihood ratio statistic G2. Subsequently, we compared risk adjustment performance of the comorbidity indices by using the Harrell c-statistic measure derived from multiple logistic regression models.

Results

Analytical results demonstrated that that comorbidity measures were significant predictors of medical expenditures and mortality of COPD patients. Specifically, in the category of diagnosis-based comorbidity indices the Elixhauser index was superior to other indices, while the RxRisk-V index was a stronger predictor in the framework of medication-based codes, for gauging both medical expenditures and in-hospital mortality by utilizing information from the index hospitalization only as well as the index and prior hospitalizations.

Conclusions

In conclusion, this work has ascertained that comorbidity indices are significant predictors of medical expenditures and mortality of COPD patients. Based on the study findings, we propose that when designing the payment schemes for patients with chronic diseases, the health authority should make adjustments in accordance with the burden of health care caused by comorbid conditions.

Occurrence of inflammatory bowel disease in patients with chronic inflammatory skin diseases: a cohort study

BACKGROUND

Several studies linked various chronic inflammatory skin diseases (CISDs) with inflammatory bowel disease (IBD) in a range of data sources with mixed conclusions.

OBJECTIVE

We compared the incidence of IBD-ulcerative colitis (UC) and Crohn's disease (CD)- in patients with a CISD versus similar persons without a CISD.

METHODS

In this cohort study using nationwide, longitudinal, commercial insurance claims data from the US, we identified adults and children who were seen by a dermatologist between 2004-2020, and diagnosed with either psoriasis, atopic dermatitis, alopecia areata, vitiligo, or hidradenitis suppurativa. Comparator patients were identified through risk-set sampling; they were eligible if they were seen by a dermatologist at least twice and not diagnosed with a CISD. Patient follow-up lasted until either IBD diagnosis, death, disenrollment, or end of data stream, whichever came first. IBD events, ulcerative colitis (UC) or Crohn's (CD), were identified via validated algorithms-hospitalization or diagnosis with endoscopic confirmation. Incidence rates were computed before and after adjustment via propensity-score (PS) decile stratification to account for IBD risk factors. Hazard ratios (HR) and 95% confidence intervals were estimated to compare the incidence of IBD in CISD versus non-CISD.

RESULTS

We identified patients with atopic dermatitis (n=123,614), psoriasis (n=83,049), alopecia areata (n=18,135), vitiligo (n=9,003) or hidradenitis suppurativa (n=6,806), and comparator patients without a CISD (n=2,376,120). During a median follow-up time of 718 days, and after applying PS adjustment for IBD risk factors, we observed increased risk of both UC (HR_UC =2.30; 1.61-3.28) and CD (HR_CD =2.70; 1.69-4.32) in patients with hidradenitis suppurativa, an increased risk of CD (HR_CD =1.23; 1.03-1.46) but not UC (HR_UC =1.01; 0.89-1.14) in psoriasis, and no increased risk of IBD in atopic dermatitis (HR_UC =1.02; 0.92-1.12, HR_CD =1.08; 0.94-1.23), alopecia areata (HR_UC =1.18; 0.89-1.56, HR_CD =1.26; 0.86-1.86) or vitiligo (HR_UC =1.14; 0.77-1.68, HR_CD =1.45; 0.87-2.41).

CONCLUSIONS

IBD was increased in patients with hidradenitis suppurativa. Crohn's disease alone was increased in patients with psoriasis. Neither ulcerative colitis nor Crohn's disease was increased in patients with atopic dermatitis, alopecia areata or vitiligo.

The comparative effectiveness of antidepressants for youths with major depressive disorder: a nationwide population-based study in Taiwan

Background:

Guidelines recommend fluoxetine as a first-line medication for youths diagnosed with major depressive disorder (MDD). However, little is known about the long-term effectiveness of different antidepressants in juveniles in the real world. This study aimed to compare the effectiveness of antidepressants in youths with MDD.

Methods:

Youths (<20 years old) with a diagnosis of MDD who were new users of antidepressants were selected from a nationwide population-based cohort in Taiwan between 1997 and 2013. We divided a total of 16,981 users (39.9% male; mean age: 16.6 years) into 10 different antidepressant groups (fluoxetine, sertraline, paroxetine, venlafaxine, citalopram, escitalopram, bupropion, fluvoxamine, mirtazapine and moclobemide). Regarding treatment outcomes (hospitalisation and medication discontinuation), Cox proportional hazards regression models were applied to estimate the hazards of such outcomes.

Results:

Compared with the youths treated with fluoxetine, the bupropion-treated group demonstrated lower rates of hospitalisation and discontinuation. Mirtazapine-treated group demonstrated a higher hospitalisation risk mainly when administered for single depressive episodes. Furthermore, patients treated with sertraline and fluvoxamine had higher discontinuation rates. Among the younger teenage subgroups (< 16 years), significantly higher rates of discontinuation were observed in those treated with sertraline, escitalopram and fluvoxamine. Among the older teenage subgroups (⩾ 16 years), bupropion was superior to fluoxetine in preventing hospitalisation and discontinuation.

Conclusion:

We concluded that bupropion might surpass fluoxetine with regard to hospitalisation prevention and drug therapy maintenance among youths with MDD, while mirtazapine users demonstrated a higher hospitalisation risk. Our findings might serve as a reference for clinicians in future studies.

Comparative Risk of Alzheimer Disease and Related Dementia Among Medicare Beneficiaries With Rheumatoid Arthritis Treated With Targeted Disease-Modifying Antirheumatic Agents

IMPORTANCE

Cytokine signaling, including tumor necrosis factor (TNF) and interleukin (IL)-6, through the Janus-kinase (JAK)-signal transducer and activator of transcription pathway, was hypothesized to attenuate the risk of Alzheimer disease and related dementia (ADRD) in the Drug Repurposing for Effective Alzheimer Medicines (DREAM) initiative based on multiomics phenotyping.

OBJECTIVE

To evaluate the association between treatment with tofacitinib, tocilizumab, or TNF inhibitors compared with abatacept and risk of incident ADRD.

DESIGN, SETTING, AND PARTICIPANTS

This cohort study was conducted among US Medicare fee-for-service patients with rheumatoid arthritis aged 65 years and older from 2007 to 2017. Patients were categorized into 3 cohorts based on initiation of tofacitinib (a JAK inhibitor), tocilizumab (an IL-6 inhibitor), or TNF inhibitors compared with a common comparator abatacept (a T-cell activation inhibitor). Analyses were conducted from August 2020 to August 2021.

MAIN OUTCOMES AND MEASURES

The main outcome was onset of ADRD based on diagnosis codes evaluated in 4 alternative analysis schemes: (1) an as-treated follow-up approach, (2) an as-started follow-up approach incorporating a 6-month induction period, (3) incorporating a 6-month symptom to diagnosis period to account for misclassification of ADRD onset, and (4) identifying ADRD through symptomatic prescriptions and diagnosis codes. Hazard ratios (HRs) with 95% CIs were calculated from Cox proportional hazard regression after adjustment for 79 preexposure characteristics through propensity score matching.

RESULTS

After 1:1 propensity score matching to patients using abatacept, a total of 22 569 propensity score-matched patient pairs, including 4224 tofacitinib pairs (mean [SD] age 72.19 [5.65] years; 6945 [82.2%] women), 6369 tocilizumab pairs (mean [SD] age 72.01 [5.46] years; 10 105 [79.4%] women), and 11 976 TNF inhibitor pairs (mean [SD] age 72.67 [5.91] years; 19 710 [82.3%] women), were assessed. Incidence rates of ADRD varied from 2 to 18 per 1000 person-years across analyses schemes. There were no statistically significant associations of ADRD with tofacitinib (analysis 1: HR, 0.90 [95% CI, 0.55-1.51]; analysis 2: HR, 0.78 [95% CI, 0.53-1.13]; analysis 3: HR, 1.29 [95% CI, 0.72-2.33]; analysis 4: HR, 0.50 [95% CI, 0.21-1.20]), tocilizumab (analysis 1: HR, 0.82 [95% CI, 0.55-1.21]; analysis 2: HR, 1.05 [95% CI, 0.81-1.35]; analysis 3: HR, 1.21 [95% CI, 0.75-1.96]; analysis 4: HR, 0.78 [95% CI, 0.44-1.39]), or TNF inhibitors (analysis 1: HR, 0.93 [95% CI, 0.72-1.20]; analysis 2: HR, 1.02 [95% CI, 0.86-1.20]; analysis 3: HR, 1.13 [95% CI, 0.86-1.48]; analysis 4: 0.90 [95% CI, 0.60-1.37]) compared with abatacept. Results from prespecified subgroup analysis by age, sex, and baseline cardiovascular disease were consistent except in patients with cardiovascular disease, for whom there was a potentially lower risk of ADRD with TNF inhibitors vs abatacept, but only in analyses 2 and 4 (analysis 1: HR, 0.76 [95% CI, 0.50-1.16]; analysis 2: HR, 0.74 [95% CI, 0.56-0.99]; analysis 3: HR, 1.03 [95% CI, 0.65-1.61]; analysis 4: HR, 0.45 [95% CI, 0.21-0.98]).

CONCLUSIONS AND RELEVANCE

This cohort study did not find any association of risk of ADRD in patients treated with tofacitinib, tocilizumab, or TNF inhibitors compared with abatacept.

Prediction Ability of Charlson, Elixhauser, and Rx-Risk Comorbidity Indices for Mortality in Patients with Hip Fracture. A Danish Population-Based Cohort Study from 2014 – 2018

Objective

Methods

Results

Conclusion

Depression and the risk of hospitalization in type 2 diabetes patients: A nested case-control study accounting for non-persistence to antidiabetic treatment

INTRODUCTION

Depression is one of the most common comorbidities of type 2 diabetes. The relationship between these two diseases seems to be bidirectional. Both conditions separately lead to significant morbidity and mortality, including hospitalization. Moreover, depression is associated with non-persistence with antidiabetic drugs.

OBJECTIVES

To measure the effect of depression on morbidity and particularly on all-cause, diabetes-related, cardiovascular-related and major cardiovascular events-related hospitalization, adjusting for non-persistence to antidiabetic drugs and other confounders.

METHODS

We performed a nested case-control study within a cohort of type 2 diabetic individuals initiating antidiabetic drugs. Using the health administrative data of the province of Quebec, Canada, we identified all-cause, diabetes-related, cardiovascular-related and major cardiovascular hospitalizations during a maximum follow-up of eight years after the initiation of antidiabetic drug treatment. A density sampling method matched all cases with up to 10 controls by age, sex, and the Elixhauser comorbidity index. The effect of depression on hospitalization was estimated using conditional logistic regressions adjusting for non-persistence to antidiabetic drug treatment and other variables.

RESULTS

We identified 41,550 all-cause hospitalized cases, of which 34,437 were related to cardiovascular (CV) diseases, 29,584 to diabetes, and 13,867 to major CV events. Depression was diagnosed in 2.51% of all-cause hospitalizations and 1.16% of matched controls. 69.11% of cases and 72.59% of controls were on metformin monotherapy. The majority (71.62% vs 75.02%, respectively) stayed on metformin monotherapy without adding or switching drugs during follow-up. Non-persistence was at similar rates (about 30%) in both groups. In the multivariable analyses, depression was associated with an increased risk for all-cause hospitalizations, with odds ratios (ORs) ranging from 2.21 (95% CI: 2.07-2.37) to 1.32 (95% CI: 1.22-1.44) according to the model adjustment (from the univariate to the fully adhjusted).

CONCLUSION

Depression increased the risk of all-cause hospitalizations among patients treated for diabetes, even after accounting for non-persistence and other potentially confounding factors. These results stress the impact of depression on diabetic patients' use of health care resources.

Charlson Comorbidity Index: A Critical Review of Clinimetric Properties

The present critical review was conducted to evaluate the clinimetric properties of the Charlson Comorbidity Index (CCI), an assessment tool designed specifically to predict long-term mortality, with regard to its reliability, concurrent validity, sensitivity, incremental and predictive validity. The original version of the CCI has been adapted for use with different sources of data, ICD-9 and ICD-10 codes. The inter-rater reliability of the CCI was found to be excellent, with extremely high agreement between self-report and medical charts. The CCI has also been shown either to have concurrent validity with a number of other prognostic scales or to result in concordant predictions. Importantly, the clinimetric sensitivity of the CCI has been demonstrated in a variety of medical conditions, with stepwise increases in the CCI associated with stepwise increases in mortality. The CCI is also characterized by the clinimetric property of incremental validity, whereby adding the CCI to other measures increases the overall predictive accuracy. It has been shown to predict long-term mortality in different clinical populations, including medical, surgical, intensive care unit (ICU), trauma, and cancer patients. It may also predict in-hospital mortality, although in some instances, such as ICU or trauma patients, the CCI did not perform as well as other instruments designed specifically for that purpose. The CCI thus appears to be clinically useful not only to provide a valid assessment of the patient's unique clinical situation, but also to demarcate major diagnostic and prognostic differences among subgroups of patients sharing the same medical diagnosis.

Differences in the association between oral corticosteroids and risk of preterm birth by data source: Reconciling the results

OBJECTIVE

To investigate causes of discrepancies in the association between early pregnancy oral corticosteroid (OCS) use and preterm birth (PTB) risk among women with rheumatoid arthritis (RA) in health care utilization [California Medicaid (Medi-Cal)] and prospective cohort (MotherToBaby Pregnancy Studies) data.

METHODS

Separately, we estimated risk ratios (RR) between OCS exposure before gestational day 140 and PTB risk in Medi-Cal (2007-2013; n=844) and MotherToBaby (2003-2014; n=528) data. We explored differences in socio-economic status, OCS dose distribution, exposure misclassification, and confounding by RA severity across the data sources.

RESULTS

PTB risk in women without OCS's was 17.3% in Medi-Cal and was 9.7% in MotherToBaby. There was no association between OCS and PTB in Medi-Cal (adjusted (a)RR: 1.00 (95% CI: 0.71, 1.42)), and a 1.85-fold (95% CI: 1.20, 2.84) increased PTB risk in MotherToBaby. When restricting each sample to women with a high school degree or less, PTB risk following no OCS exposure was 15.9% in Medi-Cal and 16.7% in MotherToBaby; aRR's were 1.16 (95% CI: 0.74, 1.80) in Medi-Cal and 0.81 (95% CI: 0.25, 2.64) in MotherToBaby. Cumulative OCS dose was higher in MotherToBaby (median: 684 mg) than Medi-Cal (median: 300 mg). OCS dose ≤300 mg was not associated with increased PTB risk. Exposure misclassification and confounding by RA severity were unlikely explanations of differences.

DISCUSSION

Higher baseline PTB risk and lower OCS dose distribution in Medi-Cal may explain the discrepancies. Studies are needed to understand the effects of autoimmune disease severity and under-treatment on PTB risk in low-income populations.

NSAID use and unnatural deaths after cancer diagnosis: a nationwide cohort study in Sweden

BACKGROUND

Cancer patients experience increased risk of death from accident and suicide. Cognitive impairment induced by cancer-related inflammation and stress-related psychiatric symptoms may be underlying mechanisms. We therefore studied the association between use of nonsteroidal anti-inflammatory drugs (NSAIDs) and risk of these outcomes.

METHODS

Following a cohort of 388,443 cancer patients diagnosed between October 2005 and December 2014 in Sweden, we ascertained dispense of aspirin or non-aspirin NSAIDs from 3 months before cancer diagnosis onward and defined the on-medication period as from date of drug dispense until the prescribed dosage was consumed. Follow-up time outside medicated periods and time from unexposed patients were defined as off-medication periods. We used Cox models to estimate hazard ratios (HRs) of death due to suicide or accident, by comparing the on-medication periods with off-medication periods.

RESULTS

In total, 29.7% of the cancer patients had low-dose aspirin dispensed and 29.1% had non-aspirin NSAIDs dispensed. Patients with aspirin use were more likely to be male than patients without aspirin use. Compared with off-medication periods, there was a 22% lower risk of accidental death (N = 651; HR 0.78, 95% confidence interval [CI]: 0.70 to 0.87) during on-medication periods with aspirin. The use of aspirin was not associated with risk of suicide (N = 59; HR 0.96, 95% CI: 0.66 to 1.39). No association was noted between use of non-aspirin NSAIDs and the risk of suicide (N = 13; HR 0.95, 95% CI: 0.42 to 2.18) or accidental death (N = 59; HR 0.92, 95% CI: 0.68 to 1.26).

CONCLUSIONS

Intake of low-dose aspirin after cancer diagnosis was associated with a lower risk of unnatural deaths among cancer patients.

Use of Palliative Care Among Commercially Insured Patients With Metastatic Cancer Between 2001 and 2016

PURPOSE

Early palliative care, concomitant with disease-directed treatments, is recommended for all patients with advanced cancer. This study assesses population-level trends in palliative care use among a large cohort of commercially insured patients with metastatic cancer, applying an expanded definition of palliative care services based on claims data.

METHODS

Using nationally representative commercial insurance claims data, we identified patients with metastatic breast, colorectal, lung, bronchus, trachea, ovarian, esophageal, pancreatic, and liver cancers and melanoma between 2001 and 2016. We assessed the annual proportions of these patients who received services specified as, or indicative of, palliative care. Using Cox proportional hazard models, we assessed whether the time from diagnosis of metastatic cancer to first encounter of palliative care differed by demographic characteristics, socioeconomic factors, or region.

RESULTS

In 2016, 36% of patients with very poor prognosis cancers received a service specified as, or indicative of, palliative care versus 18% of those with poor prognosis cancers. Being diagnosed in more recent years (2009-2016 v 2001-2008: hazard ratio [HR], 1.8; P < .001); a diagnosis of metastatic esophagus, liver, lung, or pancreatic cancer, or melanoma (v breast cancer, eg, esophagus HR, 1.89; P < .001); a greater number of comorbidities (American Hospital Formulary Service classes > 10 v 0: HR, 1.71; P < .001); and living in the Northeast (HR, 1.43; P < .001) or Midwest (v South: HR, 1.39; P < .001) were the strongest predictors of shorter time from diagnosis to palliative care.

CONCLUSION

Use of palliative care among commercially insured patients with advanced cancers has increased since 2001. However, even with an expanded definition of services specified as, or indicative of, palliative care, < 40% of patients with advanced cancers received palliative care in 2016.

Comparative effectiveness of antidepressants on geriatric depression: Real-world evidence from a population-based study

BACKGROUND

There is little real-world evidence about effectiveness of different antidepressants on geriatric depression.

METHODS

We used population-based claims data in Taiwan between 1997 and 2013 to include older patients (≥ 60 years of age) who were diagnosed with depression and started to use antidepressants. All patients were followed up until discontinuation of antidepressant use or the end of the study period. Treatment outcomes were set as the risk of switching to another antidepressant, receiving augmentation therapy, and psychiatric hospitalization. We used cox proportional hazards regression models to calculate hazard ratios with 95% confidence intervals (CIs) and adjust for several confounding factors (aHRs).

RESULTS

During the study period, a total of 207,946 elderly patients with depression received one of the following 11 antidepressants: sertraline, fluoxetine, paroxetine, escitalopram, citalopram, fluvoxamine, venlafaxine, duloxetine, moclobemide, mirtazapine, and bupropion. Compared to the patients treated with sertraline, those treated with fluvoxamine / venlafaxine had significantly but modestly higher risks of switching (aHR [95% CI]: 1.16 [1.11-1.21] / 1.10 [1.06-1.14]), augmentation (1.06 [1.02-1.10] / 1.08 [1.05-1.12]), and hospitalization (1.28 [1.03-1.58] / 1.37 [1.16-1.62]). Otherwise, the remaining 9 antidepressants yielded no consistent result in the three outcomes.

LIMITATIONS

This study is a multi-arm and active controlled trial, lacking a placebo group.

CONCLUSION

As treating geriatric depression, no individual antidepressant posed consistently better effectiveness in the outcomes of switching antidepressant, receiving augmentation, and psychiatric hospitalization than any other one, whereas clinicians should be cautious when prescribing fluvoxamine and venlafaxine.

The July Effect in Podiatric Medicine and Surgery Residency

The period when medical students begin residency in teaching hospitals throughout the United States heralds a period known in the medical community as the "July Effect." Though several sentinel studies associated this timeframe with an increase in medical errors, residencies since demystified this phenomenon within their respective specialty. This study aims to evaluate the presence of the July Effect in a podiatric medicine and surgery residency program. A retrospective chart review was conducted, comparing patient demographics and surgical outcomes including length of stay, operative time and readmission rate between the first (July, August, September) and fourth (April, May June) quarters of the academic year from 2014-2019. A total of 206 patients met the inclusion criteria, where 99 received care in the first, resident-naïve, quarter and 107 received care in the fourth, resident-experienced, quarter. No difference in patient demographics including sex, body mass index, or comorbidity index was appreciated between both quarters (p<0.05). Those patients who underwent soft tissue and bone debridements, digital, forefoot, midfoot and rearfoot amputations experienced no statistically significant difference in length of stay, operative time, or readmission rate between both quarters (p<0.05). The results of this study did not support the presence of the July Effect in our foot and ankle surgery residency. Future studies can further explore this phenomenon by examining patients admitted following traumatic injury or elective procedures. Moreover, this study shows the curriculum employed at our program provides sufficient support, guidance, and resources to limit errors attributed to the July Effect.

Risk of hospitalization and hip fracture associated with psychotropic polypharmacy in patients with dementia: A nationwide register-based study

OBJECTIVE

To investigate the association of benzodiazepines and antidepressants on the risk of hospitalization and hip fracture in patients with dementia initiating antipsychotic drug treatment.

METHODS

A register-based retrospective cohort study using data on all incident dementia cases (≥65 years) initiating antipsychotic treatment as monotherapy or in combination with benzodiazepines and/or antidepressants in Denmark from 2000 to 2015. The outcomes of interest were all-cause hospitalization and hip fracture. Cox proportional hazards models with adjustment for multiple variables were used to investigate risk of hospitalization and hip fracture within 180 days.

RESULTS

The risk of all-cause hospitalization during 180-days follow-up was significantly increased by 55% (adjusted HR: 1.55, 95% CI: 1.29-1.86, p < 0.0001), when antipsychotic use was combined with benzodiazepines, when compared to antipsychotic monotherapy. The association between the combination of antipsychotics and benzodiazepines with the risk of hip fracture did not reach statistical significance (adjusted HR: 1.50, 95% CI: 0.99-2.26, p = 0.0534).

CONCLUSIONS

The observed increased risk of all-cause hospitalization and hip fracture may indicate increased drug-related adverse events. Thus, careful and regular monitoring is needed to assess response to treatment and decrease the risk of adverse events, when antipsychotics are combined with BZDs, albeit confounding cannot be fully excluded within the current design.

Conducting Real-world Evidence Studies on the Clinical Outcomes of Diabetes Treatments

Real-world evidence (RWE), the understanding of treatment effectiveness in clinical practice generated from longitudinal patient-level data from the routine operation of the healthcare system, is thought to complement evidence on the efficacy of medications from randomized controlled trials (RCTs). RWE studies follow a structured approach. (1) A design layer decides on the study design, which is driven by the study question and refined by a medically informed target population, patient-informed outcomes, and biologically informed effect windows. Imagining the randomized trial we would ideally perform before designing an RWE study in its likeness reduces bias; the new-user active comparator cohort design has proven useful in many RWE studies of diabetes treatments. (2) A measurement layer transforms the longitudinal patient-level data stream into variables that identify the study population, the pre-exposure patient characteristics, the treatment, and the treatment-emergent outcomes. Working with secondary data increases the measurement complexity compared to primary data collection that we find in most RCTs. (3) An analysis layer focuses on the causal treatment effect estimation. Propensity score analyses have gained in popularity to minimize confounding in healthcare database analyses. Well-understood investigator errors, like immortal time bias, adjustment for causal intermediates, or reverse causation, should be avoided. To increase reproducibility of RWE findings, studies require full implementation transparency. This article integrates state-of-the-art knowledge on how to conduct and review RWE studies on diabetes treatments to maximize study validity and ultimately increased confidence in RWE-based decision making.

Current Trends in Readmission Prediction: An Overview of Approaches

Hospital readmission shortly after discharge threatens the quality of patient care and leads to increased medical care costs. In the United States, hospitals with high readmission rates are subject to federal financial penalties. This concern calls for incentives for healthcare facilities to reduce their readmission rates by predicting patients who are at high risk of readmission. Conventional practices involve the use of rule-based assessment scores and traditional statistical methods, such as logistic regression, in developing risk prediction models. The recent advancements in machine learning driven by improved computing power and sophisticated algorithms have the potential to produce highly accurate predictions. However, the value of such models could be overrated. Meanwhile, the use of other flexible models that leverage simple algorithms offer great transparency in terms of feature interpretation, which is beneficial in clinical settings. This work presents an overview of the current trends in risk prediction models developed in the field of readmission. The various techniques adopted by researchers in recent years are described, and the topic of whether complex models outperform simple ones in readmission risk stratification is investigated.

Natural language processing for the assessment of cardiovascular disease comorbidities: The cardio-Canary comorbidity project

Objective: Accurate ascertainment of comorbidities is paramount in clinical research. While manual adjudication is labor‐intensive and expensive, the adoption of electronic health records enables computational analysis of free‐text documentation using natural language processing (NLP) tools.

Hypothesis: We sought to develop highly accurate NLP modules to assess for the presence of five key cardiovascular comorbidities in a large electronic health record system.

Methods: One‐thousand clinical notes were randomly selected from a cardiovascular registry at Mass General Brigham. Trained physicians manually adjudicated these notes for the following five diagnostic comorbidities: hypertension, dyslipidemia, diabetes, coronary artery disease, and stroke/transient ischemic attack. Using the open‐source Canary NLP system, five separate NLP modules were designed based on 800 “training‐set” notes and validated on 200 “test‐set” notes.

Results: Across the five NLP modules, the sentence‐level and note‐level sensitivity, specificity, and positive predictive value was always greater than 85% and was most often greater than 90%. Accuracy tended to be highest for conditions with greater diagnostic clarity (e.g. diabetes and hypertension) and slightly lower for conditions whose greater diagnostic challenges (e.g. myocardial infarction and embolic stroke) may lead to less definitive documentation.

Conclusion: We designed five open‐source and highly accurate NLP modules that can be used to assess for the presence of important cardiovascular comorbidities in free‐text health records. These modules have been placed in the public domain and can be used for clinical research, trial recruitment and population management at any institution as well as serve as the basis for further development of cardiovascular NLP tools.

Muscle volume and intramuscular fat of the tongue evaluated with MRI predict malnutrition in people living with dementia: a five-year follow-up study

Malnutrition is highly prevalent in older persons with dementia. Therefore, strong predictors of malnutrition in this population are crucial to initiating early interventions. This study evaluates the association between the probability of having malnutrition with the muscle volume and intramuscular fat (iMAT) of the masseter and the tongue in magnetic resonance images (MRI) of community-dwelling older persons diagnosed with mild dementia followed for five years. This is a longitudinal study conducted in the western part of Norway. Muscle volume and iMAT of the tongue and masseter were computed from structural head MRIs obtained from 65 participants of The Dementia Study of Western Norway (DemVest) using Slice-O-Matic software for segmentation. Malnutrition was assessed using the glim index. Linear mix models were conducted. Having malnutrition at baseline was associated with lower muscle volume (OR 0.60 SE 0.20 p=0.010) and higher iMAT (OR 3.31 SE 0.46 p=0.010) in the tongue. At five years follow-up, those with lower muscle volume (OR 0.55, SE 0.20 p=0.002) and higher iMAT (OR 2.52, SE 0.40 p=0.022) in the tongue had a higher probability of presenting malnutrition. The masseter iMAT and volume were not associated with malnutrition in any of the adjusted models.In people diagnosed with mild dementia, tongue muscle volume and iMAT were associated with baseline malnutrition and the probability of developing malnutrition in a 5-year trajectory. In the masseter, there were no significant associations after adjustments.

Risk Factors Associated with 90-day Readmissions Following Odontoid Fractures: A Nationwide Readmissions Database Study

STUDY DESIGN

Nationwide Readmissions Database Study.

OBJECTIVE

The aim of this study was to investigate readmission rates and factors related to readmission after surgical and nonsurgical management of odontoid fractures.

SUMMARY OF BACKGROUND DATA

Management of odontoid fractures, which are the most common isolated spine fracture in the elderly, continues to be debated. The choice between surgical or nonsurgical treatment has been reported to impact mortality and might influence readmission rates. Hospital readmissions represent a large financial burden upon our healthcare system. Factors surrounding hospital readmissions would benefit from a better understanding of their associated causes to lower health care costs.

METHODS

A retrospective study was performed using the 2016 Healthcare Utilization Project (HCUP) Nationwide Readmission Database (NRD). Demographic information and factors associated with readmission were collected. Readmission rates, complications, length of hospital stay were collected. Patients treated operatively, nonoperatively, and patients who were readmitted or not readmitted were compared. Statistical analysis was performed using open source software SciPy (Python v1.3.0) for all analyses.

RESULTS

We identified 2921 patients who presented with Type II dens fractures from January 1, 2016 to September 30, 2016, 555 of which underwent surgical intervention. The readmission rate in patients who underwent surgery was 16.4% (91/555) and 29.4% (696/2366) in the nonoperative group. Hospital costs for readmitted and nonreadmitted patients were $353,704 and $174,922, and $197,099 and $80,715 for nonoperatively managed patients, respectively. Medicaid and Medicare patients had the highest readmission rate in both groups. Charlson and Elixhauser comorbidity indices were significantly higher in patients who were readmitted (P < 0.0001).

CONCLUSION

We report an overall 90-day readmission rate of 16.4% and 29.4%, in operative and nonoperative management of type II odontoid fractures, respectively. In the face of a rising incidence of this fracture in the elderly population, an understanding of the comorbidities and age-related demographics associated with 90-day readmissions following both surgical and nonsurgical treatment are critical.Level of Evidence: 3.

Real-world evidence of bariatric surgery and cardiovascular benefits using electronic health records data: A lesson in bias

AIM

To reproduce and correct studies on bariatric surgery and the reduction in major adverse cardiovascular events (MACE) among patients with obesity and type 2 diabetes (T2D).

METHODS

We used electronic healthcare records (EHR) from in and outpatient facilities around the United States to identify a cohort of patients with T2D, aged 18 to 80 years and with a body mass index (BMI) of 30 kg/m² or higher undergoing bariatric surgery. We compared against hip/knee arthroplasty to establish an active comparison group that reduced bias from differential information and confounding. The main outcome was six-point MACE. Pre-exposure characteristics were adjusted in propensity score (PS) models with 1:2 matching plus high-dimensional PS matching.

RESULTS

After a range of exclusions, the final cohort included 344 bariatric surgery patients (65% female; mean age 58 years) and 551 PS-matched patients undergoing arthroplasty (65% female; 59 years). Median follow-up was 2.5 years in both groups. Bariatric surgery patients showed a sustained 20% weight reduction and an HbA1c reduction by 1% point. We found no benefits of bariatric surgery for six-point MACE (HR = 0.99; 95% CI 0.76-1.30). We observed known increases in risks for vitamin B₁₂ deficiency anaemia (HR = 3.06; 1.10-8.49) and cholelithiasis (HR = 1.72; 0.94-3.13).

CONCLUSIONS

This real-world evidence study found reductions in HbA1c and BMI following bariatric surgery similar to trials, and no meaningful cardiovascular benefit compatible with the underpowered trials but in contrast to earlier EHR studies. We showed how information bias typical in EHR analyses and confounding may cause substantial bias.

A narrative review of using prescription drug databases for comorbidity adjustment: A less effective remedy or a prescription for improved model fit?

BACKGROUND

The use of claims data for identifying comorbid conditions in patients for research purposes has been widely explored. Traditional measures of comorbid adjustment included diagnostic data (e.g., ICD-9-CM or ICD-10-CM codes), with the Charlson and Elixhauser methodology being the two most common approaches. Prescription data has also been explored for use in comorbidity adjustment, however early methodologies were disappointing when compared to diagnostic measures.

OBJECTIVE

The objective of this methodological review is to compare results from newer studies using prescription-based data with more traditional diagnostic measures.

METHODS

A review of studies found on PubMed, Medline, Embase or CINAHL published between January 1990 and December 2020 using prescription data for comorbidity adjustment. A total of 50 studies using prescription drug measures for comorbidity adjustment were found.

CONCLUSIONS

Newer prescription-based measures show promise fitting models, as measured by predictive ability, for research, especially when the primary outcomes are utilization or drug expenditure rather than diagnostic measures. More traditional diagnostic-based measures still appear most appropriate if the primary outcome is mortality or inpatient readmissions.

Healthcare resource utilisation and sickness absence in newly diagnosed multiple myeloma patients who did not undergo autologous stem cell transplantation: Trends in Sweden with the changing treatment landscape

OBJECTIVES

The introduction of novel drugs has significantly improved outcomes for multiple myeloma (MM) patients. This study describes survival, healthcare resource utilisation and sickness absence in association with the changing MM treatment landscape over time, focussing on patients who did not undergo autologous stem cell transplantation (ASCT).

METHODS

Population-based, retrospective registry study in Sweden, where 7012 non-ASCT patients diagnosed between 2001 and 2015 were stratified into diagnosis periods 2001-2005 (n = 2053), 2006-2010 (n = 2372) and 2011-2015 (n = 2587).

RESULTS

Median survival increased from 2.5 to 3.4 years from 2001-2005 to 2011-2015. During the first 3 years of follow-up, patients diagnosed during 2011-2015 spent 29% and 12% less time in health care (55 days; inpatient admissions and outpatient visits) than patients diagnosed during 2001-2005 (78 days) and 2006-2010 (63 days), respectively. This was associated with less inpatient and more outpatient healthcare usage. Average 3-year sickness absence (362 days) was 31% and 12% less than for patients diagnosed during 2001-2005 (522 days) and 2006-2010 (410 days), respectively.

CONCLUSIONS

These findings of improved survival, reduced healthcare needs and greater productivity in non-ASCT MM patients with access to improved treatment practices and novel drugs provide important real-world cost-benefit insights for the continued development and introduction of treatments for MM.

Dealing With Treatment-Confounder Feedback and Sparse Follow-up in Longitudinal Studies: Application of a Marginal Structural Model in a Multiple Sclerosis Cohort

The beta-interferons are widely prescribed platform therapies for patients with multiple sclerosis (MS). We accessed a cohort of patients with relapsing-onset MS from British Columbia, Canada (1995-2013), to examine the potential survival advantage associated with beta-interferon exposure using a marginal structural model. Accounting for potential treatment-confounder feedback between comorbidity, MS disease progression, and beta-interferon exposure, we found an association between beta-interferon exposure of at least 6 contiguous months and improved survival (hazard ratio (HR) = 0.63, 95% confidence interval 0.47, 0.86). We also assessed potential effect modifications by sex, baseline age, or baseline disease duration, and found these factors to be important effect modifiers. Sparse follow-up due to variability in patient contact with the health system is one of the biggest challenges in longitudinal analyses. We considered several single-level and multilevel multiple imputation approaches to deal with sparse follow-up and disease progression information; both types of approach produced similar estimates. Compared to ad hoc imputation approaches, such as linear interpolation (HR = 0.63), and last observation carried forward (HR = 0.65), all multiple imputation approaches produced a smaller hazard ratio (HR = 0.53), although the direction of effect and conclusions drawn concerning the survival advantage remained the same.

Development and Validation of a Pediatric Comorbidity Index

Comorbidity scores are widely used to help address confounding bias in nonrandomized studies conducted within health-care databases, but existing scores were developed to predict all-cause mortality in adults and might not be appropriate for use in pediatric studies. We developed and validated a pediatric comorbidity index, using health-care utilization data from the tenth revision of the International Classification of Diseases. Within the MarketScan database of US commercial claims data, pediatric patients (aged ≤18 years) continuously enrolled between October 1, 2015, and September 30, 2017, were identified. Logistic regression was used to predict the 1-year risk of hospitalization based on 27 predefined conditions and empirically identified conditions derived from the most prevalent diagnoses among patients with the outcome. A single numerical index was created by assigning weights to each condition based on its β coefficient. We conducted internal validation of the index and compared its performance with existing adult scores. The pediatric comorbidity index consisted of 24 conditions and achieved a C statistic of 0.718 (95% confidence interval (CI): 0.714, 0.723). The index outperformed existing adult scores in a pediatric population (C statistics ranging from 0.522 to 0.640). The pediatric comorbidity index provides a summary measure of disease burden and can be used for risk adjustment in epidemiologic studies of pediatric patients.

Interaction between Omeprazole and Gliclazide in Relation to CYP2C19 Phenotype

The antidiabetic drug gliclazide is partly metabolized by CYP2C19, the main enzyme involved in omeprazole metabolism. The aim of the study was to explore the interaction between omeprazole and gliclazide in relation to CYP2C19 phenotype using physiologically based pharmacokinetic (PBPK) modeling approach. Developed PBPK models were verified using in vivo pharmacokinetic profiles obtained from a clinical trial on omeprazole-gliclazide interaction in healthy volunteers, CYP2C19 normal/rapid/ultrarapid metabolizers (NM/RM/UM). In addition, the association of omeprazole cotreatment with gliclazide-induced hypoglycemia was explored in 267 patients with type 2 diabetes (T2D) from the GoDARTS cohort, Scotland. The PBPK simulations predicted 1.4–1.6-fold higher gliclazide area under the curve (AUC) after 5-day treatment with 20 mg omeprazole in all CYP2C19 phenotype groups except in poor metabolizers. The predicted gliclazide AUC increased 2.1 and 2.5-fold in intermediate metabolizers, and 2.6- and 3.8-fold in NM/RM/UM group, after simulated 20-day dosing with 40 mg omeprazole once and twice daily, respectively. The predicted results were corroborated by findings in patients with T2D which demonstrated 3.3-fold higher odds of severe gliclazide-induced hypoglycemia in NM/RM/UM patients concomitantly treated with omeprazole. Our results indicate that omeprazole may increase exposure to gliclazide and thus increase the risk of gliclazide-associated hypoglycemia in the majority of patients.

More than two-dozen prescription drug-based risk scores are available for risk adjustment: A systematic review

OBJECTIVE

While several prescription drug-based risk indices have been developed, their design, performance, and application has not previously been synthesized.

STUDY DESIGN AND SETTING

We searched Ovid MEDLINE, CINAHL and Embase from inception through March 3, 2020 and included studies that developed or updated a prescription drug-based risk index. Two reviewers independently performed screening and extracted information on data source, study population, cohort sizes, outcomes, study methodology and performance. Predictive performance was evaluated using C statistics for binary outcomes and R² for continuous outcomes. The PROSPERO ID for this review is CRD42020165498.

RESULTS

Of 19,112 articles that were retrieved, 124 were full-text screened and 25 were included, each of which represented a de novo or updated drug-based index. The indices were customized to varied age groups and clinical populations and most commonly evaluated outcomes including mortality (36%), hospitalization (24%) and healthcare costs (24%). C statistics ranged from 0.62 to 0.92 for mortality and 0.59 to 0.72 for hospitalization, while adjusted R² for healthcare costs ranged from 0.06 to 0.62. Seven of the 25 risk indices included used global drug classification algorithms.

CONCLUSIONS

More than two-dozen prescription drug-based risk indices have been developed and they differ significantly in design, performance and application.

Pharmacosafety of fluoroquinolone and macrolide antibiotics in the clinical care of patients with myasthenia gravis

INTRODUCTION/AIMS

Anecdotal case reports have suggested a potential association of fluoroquinolones and macrolides with myasthenia gravis (MG) exacerbation, prompting warnings against the use of these drugs in this population. However, large-scale and reliable population-based data that demonstrate this association are lacking. This study aims to examine the association between outpatient treatment with fluoroquinolones or macrolides and MG-related hospitalization.

METHODS

A retrospective cohort study consisting of adult MG patients was conducted using a large de-identified healthcare claims database. Antibiotic prescription claims were identified, and MG-related hospitalizations were assessed at 15, 30, and 90 days after the date of prescription. We used mixed effects survival regression with log-logistic distribution and independent covariance matrix to estimate odds ratios (ORs) of hospitalization for each potentially exacerbating antibiotic using beta-lactam as the reference and adjusting for covariates.

RESULTS

Among 1556 MG patients receiving 894 fluoroquinolone prescriptions, 729 macrolide prescriptions, and 1608 beta-lactam prescriptions during the study period, there was no difference in 15, 30, or 90-day odds of MG-related hospitalization between fluoroquinolone or macrolide users compared to prescribed beta-lactams. However, estimates were higher for fluoroquinolones than macrolides, even after covariate adjustment (adjusted OR [aOR] 4.60, 95% confidence interval [CI] 0.55-38.57 for fluoroquinolones and OR 0.56, 95% CI 0.32-0.97 for macrolides, respectively, at 15 days).

DISCUSSION

Fluoroquinolone and macrolide antibiotics are prescribed frequently to patients with MG. While statistical imprecision precludes a definitive conclusion, elevated ORs for fluoroquinolones raise the possibility of an underpowered association that merits further investigation.

In-hospital myocardial infarction and adherence to evidence-based drug therapies: a real-world evaluation

OBJECTIVES

This study aimed to measure adherence to chronic polytherapy following an acute myocardial infarction (AMI) and to find out associations between adherence and the setting of AMI onset (in vs out of hospital) as well as other determinants.

DESIGN

Retrospective follow-up study.

SETTING

Population living in the Lazio Region, Italy.

PARTICIPANTS

This study included 25 779 hospitalised patients with a first diagnosis of AMI in 2012-2016, after the exclusion of those with hospital admission for AMI or related causes in the previous 5 years.

PRIMARY AND SECONDARY OUTCOME MEASURES

Patients were classified as in-hospital AMI (IH-AMI) or out of hospital AMI (OH-AMI) according to present-on-admission codes. Adherence was measured based on prescription claims during a 6-month follow-up after hospital discharge, using medication possession ratio (MPR). Adherence to chronic polytherapy was defined as MPR ≥75% to at least 3 of the following medications: antithrombotics, betablockers, ACE inhibitors/angiotensin receptor blockers and statins.

RESULTS

Among the entire cohort, 1 044 (4%) patients suffered IH-AMI. Overall, 15 440 (60%) patients were deemed adherent to chronic polytherapy. Female gender, older age, mental disorders, renal disease, asthma and ongoing concomitant treatments were factors associated with poor adherence. By contrast, patients with more severe AMI and those already taking evidence-based (E-B) drugs were more likely to be adherent. A strong association between the setting of AMI onset and adherence was observed: IH-AMI patients were 46% less likely to be adherent to E-B medications during their 6-month follow-up as compared with OH-AMI patients (OR 0.54; 95% CI 0.47 to 0.62; p<0.001).

CONCLUSION

Pharmacotherapy is not consistent with clinical guidelines, especially for IH-AMI patients. Our findings provide evidence on a previously unidentified groups of patients at risk for poor adherence, who might benefit from greater medical attention and dedicated healthcare interventions.

Medication adherence in multiple sclerosis as a potential model for other chronic diseases: a population-based cohort study

OBJECTIVE

To determine whether better medication adherence in multiple sclerosis (MS) might be due to specialised disease-modifying drug (DMD) support programmes by: (1) establishing higher adherence in MS than in other chronic diseases and (2) determining if higher adherence is associated with patient-specific or treatment-specific factors.

DESIGN

Retrospective cohort study with data from 1 January 1996 to 31 December 2015.

SETTING

Population-based health administrative data from three Canadian provinces.

PARTICIPANTS

Individual cohorts were created using validated case definitions for MS, epilepsy, Parkinson's disease (PD) and rheumatoid arthritis (RA). Subjects were included if they received ≥1 dispensation for a disease-related drug between 1 January 1997 and 31 December 2014.

MAIN OUTCOME MEASURES

Proportion of subjects with optimal adherence (≥80%) measured by the medication possession ratio 1 year after the index date (first dispensation of disease-related drug).

RESULTS

126 478 subjects were included in the primary analysis (MS, n=6271; epilepsy, n=55 739; PD, n=21 304; RA, n=43 164). Subjects with epilepsy (adjusted OR, aOR 0.29; 95% CI 0.19 to 0.45), PD (aOR 0.42; 95% CI 0.29 to 0.63) or RA (aOR 0.26; 95% CI 0.19 to 0.35) were less likely to have optimal 1-year adherence compared with subjects with MS. Within the MS cohort, adherence was higher for DMD than for chronic-use non-MS medications, and no consistent patient-related predictors of adherence were observed across all four non-MS medication classes, including having optimal adherence to DMD.

CONCLUSIONS

Subjects with MS were significantly more likely to have optimal 1-year adherence than subjects with epilepsy, RA and PD, and optimal adherence appears related to treatment-specific factors rather than patient-related factors. This supports the hypothesis that higher adherence to the MS DMDs could be due to the specialised support programmes; these programmes may serve as a model for use in other chronic conditions.

Identifying subgroups and risk among frequent emergency department users in British Columbia

Objective: Frequent emergency department (ED) users are heterogeneous. We aimed to identify subgroups and assess their mortality. Methods: We identified patients ≥18 years with ≥1 ED visit in British Columbia from April 1, 2012 to March 31, 2015, and linked to hospitalization, physician billing, prescription, and mortality data. Frequent users were the top 10% of patients by ED visits. We employed cluster analysis to identify frequent user subgroups. We assessed 365-day mortality using Kaplan-Meier curves and conducted Cox regressions to assess mortality risk factors within subgroups. Results: We identified 4 subgroups. Subgroup 1 ("Elderly") had median age 77 years (interquartile range [IQR]: 66-85), 5 visits/year (IQR: 4-6), median 8 prescription medications (IQR: 5-11), and 24.7% mortality. Subgroup 2 ("Mental Health and Alcohol Use") had median age 48 years (IQR: 34-61), 13 visits/year (IQR: 10-16), and 12.3% mortality. They made a median 31 general practitioner visits (IQR: 19-51); however, only 23.7% received a majority of services from 1 primary care physician. Subgroup 3 ("Young Mental Health") had median age 39 years (IQR: 28-51), 5 visits/year (IQR: 4-6), and 2.2% mortality. Subgroup 4 ("Short-term") had median age 50 years (IQR: 34-65), 4 visits/year (IQR: 4-5) regularly spaced over a short term, and 1.4% mortality. Male sex (all subgroups), long-term care ("Mental Health and Alcohol Use;" "Young Mental Health"), and rural residence ("Elderly" in long-term care; "Young Mental Health") were associated with increased mortality. Conclusions: Our results identify frequent user subgroups with varying mortality. Future research should explore subgroups' unmet needs and tailor interventions toward them.

Oral Glucocorticoids and Incident Treatment of Diabetes Mellitus, Hypertension, and Venous Thromboembolism in Children

Rates of incident treatment were quantified in this study for diabetes mellitus, hypertension, and venous thromboembolism (VTE) associated with oral glucocorticoid exposure in children aged 1-18 years. The retrospective cohort included more than 930,000 children diagnosed with autoimmune diseases (namely, inflammatory bowel disease, juvenile idiopathic arthritis, or psoriasis) or a nonimmune comparator condition (attention-deficit/hyperactivity disorder) identified using US Medicaid claims (2000-2010). Associations of glucocorticoid dose per age- and sex-imputed weight with incident treated diabetes, hypertension, and VTE were estimated using Cox regression models. Crude rates were lowest for VTE (unexposed: 0.5/million person-days (95% confidence interval (CI): 0.4, 0.6); currently exposed: 15.6/million person-days (95% CI: 11.8, 20.1)) and highest for hypertension (unexposed: 6.7/million person-days (95% CI: 6.5, 7.0); currently exposed: 74.4/million person-days (95% CI: 65.7, 83.9)). Absolute rates for all outcomes were higher in unexposed and exposed children with autoimmune diseases compared with those with attention-deficit/hyperactivity disorder. Strong dose-dependent relationships were found between current glucocorticoid exposure and all outcomes (adjusted hazard ratios for high-dose glucocorticoids: for diabetes mellitus, 5.93 (95% CI: 3.94, 8.91); for hypertension, 19.13 (95% CI: 15.43, 23.73); for VTE, 16.16 (95% CI: 8.94, 29.22)). These results suggest strong relative risks, but low absolute risks, of newly treated VTE, diabetes, and especially hypertension in children taking high-dose oral glucocorticoids.