Healthcare Utilization and Direct Costs Prior to Subspecialty Care in Children with Chronic Pain Compared with Other Chronic Childhood Diseases: A Cohort Study

OBJECTIVES

To understand the burden associated with pediatric chronic pain (CP) on the healthcare system compared with other costly chronic diseases prior to subspecialty care.

STUDY DESIGN

In this retrospective cohort study, we assessed all-cause healthcare utilization and direct healthcare costs associated with pediatric CP (n=91) compared with juvenile-onset arthritis (JA, n=135), inflammatory bowel disease (IBD, n=90), type 1 or type 2 diabetes (T1D, n=475, T2D, n=289), anxiety (n=7,193) and controls (n=273) two and five years prior to patients entering subspecialty care in Manitoba, Canada. Linked data from physician encounters, emergency department visits, hospitalizations, and prescriptions were extracted from administrative databases. Differences in healthcare utilization and direct healthcare costs associated with CP versus the other conditions were tested using negative binomial and zero-inflated negative binomial regression models, respectively.

RESULTS

After adjustment for age at diagnosis, sex, location of residence, and socioeconomic status, CP continued to be associated with the highest number of consulted physicians and subspecialists and the highest number of physician billings compared with all other conditions (p<0.01, respectively). CP was significantly associated with higher physician costs than JA, IBD, T1D, T2D, or controls (p<0.01, respectively); anxiety was associated with the highest physician and prescription costs among all cohorts (p<0.01, respectively).

CONCLUSION

Compared with chronic inflammatory and endocrinologic conditions, pediatric CP and anxiety were associated with substantial burden on the healthcare system prior to subspecialty care, suggesting a need to assess gaps and resources in the management of CP and mental health conditions in the primary care setting.

Sex-Based Disparities in Acute Myocardial Infarction Treatment Patterns and Outcomes in Older Adults Hospitalized Across 6 High-Income Countries: An Analysis From the International Health Systems Research Collaborative

BACKGROUND

Sex differences in acute myocardial infarction treatment and outcomes are well documented, but it is unclear whether differences are consistent across countries. The objective of this study was to investigate the epidemiology, use of interventional procedures, and outcomes for older females and males hospitalized with ST-segment-elevation myocardial infarction (STEMI) and non-ST-segment-elevation myocardial infarction (NSTEMI) in 6 diverse countries.

METHODS

We conducted a serial cross-sectional cohort study of 1 508 205 adults aged ≥66 years hospitalized with STEMI and NSTEMI between 2011 and 2018 in the United States, Canada, England, the Netherlands, Taiwan, and Israel using administrative data. We compared females and males within each country with respect to age-standardized hospitalization rates, rates of cardiac catheterization, percutaneous coronary intervention, and coronary artery bypass graft surgery within 90 days of hospitalization, and 30-day age- and comorbidity-adjusted mortality.

RESULTS

Hospitalization rates for STEMI and NSTEMI decreased between 2011 and 2018 in all countries, although the hospitalization rate ratio (rate in males/rate in females) increased in virtually all countries (eg, US STEMI ratio, 1.58:1 in 2011 and 1.73:1 in 2018; Israel NSTEMI ratio, 1.71:1 in 2011 and 2.11:1 in 2018). Rates of cardiac catheterization, percutaneous coronary intervention, and coronary artery bypass graft surgery were lower for females than males for STEMI in all countries and years (eg, US cardiac catheterization in 2018, 88.6% for females versus 91.5% for males; Israel percutaneous coronary intervention in 2018, 76.7% for females versus 84.8% for males) with similar findings for NSTEMI. Adjusted mortality for STEMI in 2018 was higher for females than males in 5 countries (the United States, Canada, the Netherlands, Israel, and Taiwan) but lower for females than males in 5 countries for NSTEMI.

CONCLUSIONS

We observed a larger decline in acute myocardial infarction hospitalizations for females than males between 2011 and 2018. Females were less likely to receive cardiac interventions and had higher mortality after STEMI. Sex disparities seem to transcend borders, raising questions about the underlying causes and remedies.

Variation in care for patients presenting with hip fracture in six high-income countries: A cross-sectional cohort study

BACKGROUND

Hip fractures are costly and common in older adults, but there is limited understanding of how treatment patterns and outcomes might differ between countries.

METHODS

We performed a retrospective serial cross-sectional cohort study of adults aged ≥66 years hospitalized with hip fracture between 2011 and 2018 in the US, Canada, England, the Netherlands, Taiwan, and Israel using population-representative administrative data. We examined mortality, hip fracture treatment approaches (total hip arthroplasty [THA], hemiarthroplasty [HA], internal fixation [IF], and nonoperative), and health system performance measures, including hospital length of stay (LOS), 30-day readmission rates, and time-to-surgery.

RESULTS

The total number of hip fracture admissions between 2011 and 2018 ranged from 23,941 in Israel to 1,219,696 in the US. In 2018, 30-day mortality varied from 3% (16% at 1 year) in Taiwan to 10% (27%) in the Netherlands. With regards to processes of care, the proportion of hip fractures treated with HA (range 23%-45%) and THA (0.2%-10%) differed widely across countries. For example, in 2018, THA was used to treat approximately 9% of patients in England and Israel but less than 1% in Taiwan. Overall, IF was the most common surgery performed in all countries (40%-60% of patients). IF was used in approximately 60% of patients in the US and Israel, but only 40% in England. In 2018, rates of nonoperative management ranged from 5% of patients in Taiwan to nearly 10% in England. Mean hospital LOS in 2018 ranged from 6.4 days (US) to 18.7 days (England). The 30-day readmission rate in 2018 ranged from 8% (in Canada and the Netherlands) to nearly 18% in England. The mean days to surgery in 2018 ranged from 0.5 days (Israel) to 1.6 days (Canada).

CONCLUSIONS

We observed substantial between-country variation in mortality, surgical approaches, and health system performance measures. These findings underscore the need for further research to inform evidence-based surgical approaches.

Fractures and long-term mortality in cancer patients: a population-based cohort study

We assessed post-fracture mortality in a population-based cohort of 122,045 individuals with cancers. Major fractures (hip, vertebrae, humerus, and forearm) were associated with early and long-term increased all-cause mortality.

INTRODUCTION

Currently, there are no population-based data among cancer patients on post-fracture mortality risk across a broad range of cancer diagnoses. Our objective was to estimate the association of fracture with mortality in cancer survivors.

METHODS

Using Manitoba Cancer Registry data from the province of Manitoba, Canada, we identified all women and men with cancer diagnosed between January 1, 1987, and March 31, 2014. We then linked cancer data to provincial healthcare administrative data and ascertained fractures after cancer diagnosis and mortality to March 31, 2015. Hazard ratios for all-cause mortality in those with versus without fracture were estimated from time-dependent Cox proportional hazards models adjusted for multiple covariates.

RESULTS

The study cohort consisted of 122,045 cancer patients (median age 68 years, IQR 58-77, 49.2% female). During the median follow-up of 5.8 years from cancer diagnosis, we ascertained 7120 (5.8%) major fractures. All fracture sites, except for the forearm, were associated with increased mortality risk, even after multivariable adjustment. Excess mortality risk associated with a major fracture was greatest in the first year after fracture (HR 2.42, 95% CI 2.30-2.54) and remained significant > 5 years after fracture (HR 1.60, 95% CI 1.50-1.70) and for fractures occurring > 10 years after cancer diagnosis (HR 1.93, 95% CI 1.79-2.07).

CONCLUSION

Fractures among cancer patients are associated with increased all-cause mortality. This excess risk is greatest in the first year and persists more than 5 years post-fracture; increased risk is also noted for fractures occurring up to and beyond 10 years after cancer diagnosis.

Identification of Early Onset Dementia in Population-Based Health Administrative Data: A Validation Study Using Primary Care Electronic Medical Records

BACKGROUND

Early onset dementia (EOD) occurs when symptoms of dementia begin between 45 to 64 years of age.

OBJECTIVE

We developed and validated health administrative data algorithms for EOD and compared demographic characteristics and presence of comorbid conditions amongst adults with EOD, late onset dementia (LOD) and adults with no dementia in Ontario, Canada.

METHODS

Patients aged 45 to 64 years identified as having EOD in their primary care electronic medical records had their records linked to provincial health administrative data. We compared several combinations of physician's claims, hospitalizations, emergency department visits and prescriptions. Age-standardized incidence and prevalence rates of EOD were estimated from 1996 to 2016.

RESULTS

The prevalence of EOD for adults aged 45 to 64 years in our primary care reference cohort was 0.12% . An algorithm of ≥1 hospitalization or ≥3 physician claims at least 30 days apart in a two-year period or ≥1 dementia medication had a sensitivity of 72.9% (64.5-81.3), specificity of 99.7% (99.7-99.8), positive predictive value (PPV) of 23.7% (19.1-28.3), and negative predictive value of 100.0% . Multivariate logistic regression found adults with EOD had increased odds ratios for several health conditions compared to LOD and no dementia populations. From 1996 to 2016, the age-adjusted incidence rate increased slightly (0.055 to 0.061 per 100 population) and the age-adjusted prevalence rate increased three-fold (0.11 to 0.32 per 100 population).

CONCLUSION

While we developed a health administrative data algorithm for EOD with a reasonable sensitivity, its low PPV limits its ability to be used for population surveillance.

A Synthesis of Algorithms for Multi-Jurisdiction Research in Canada

ObjectivesValidation of algorithms to identify health conditions (e.g., diabetes) or service use (e.g., high-cost users) in administrative data is time-consuming and expensive. Many algorithms are only assessed in a single jurisdiction, which may limit generalizability. Our study described the characteristics of multi-jurisdiction algorithms from a Canadian algorithm repository. ApproachWe summarized algorithms captured in the open-access Algorithms Inventory developed by Health Data Research Network (HDRN) Canada. This inventory contains published algorithms identified through a series of systematic reviews of peer-reviewed research. Algorithms included in the inventory were validated or assessed for feasibility of implementation in two or more provinces/territories; they encompass measures of population health, health service use, and determinants of health. Descriptive statistics were used to characterize the study data on such features as year and discipline of the study journal, algorithm topic area, jurisdictions included in the study, validation source data, and algorithm elements (i.e., diagnosis codes). ResultsThe HDRN Canada Algorithms Inventory currently contains 166 algorithms from 63 published articles. The majority of articles were published in 2010 or later (89%) and more than half (56%) of the articles were found in journals with a clinical focus. Feasibility studies (79%) were conducted more often than validation studies (21%). Most algorithms used data from the provinces of British Columbia, Manitoba, Ontario, and Nova Scotia. The majority of algorithms (72%) measured population health concepts, such as chronic physical health conditions (63%; e.g., hypertension) and mental health conditions (14%; e.g., depression). Algorithms about the determinants of health (17%) mostly focused on measures of socioeconomic status (37%) derived from census data. Multi-jurisdiction algorithms about health service use were least common (11%). ConclusionThis synthesis revealed few Canadian multi-jurisdiction validation studies have been conducted and not all provinces/territories are equally represented. New validation studies, particularly about health service use and determinants of health, will increase the consistency and accuracy of Canadian research. Reusing published algorithms from this inventory will facilitate research reproducibility.

Cancer patients with fractures are rarely assessed or treated for osteoporosis: a population-based study

Among 4238 cancer and 16,418 cancer-free individuals with incident major non-traumatic fractures (hip, clinical vertebral, forearm, humerus), post-fracture osteoporosis care was equally poor for both groups, whether assessed from bone mineral density (BMD) testing, initiation of osteoporosis therapy or either intervention (BMD testing and/or osteoporosis therapy).

INTRODUCTION

Most individuals sustaining a fracture do not undergo evaluation and/or treatment for osteoporosis. Cancer survivors are at increased risk for osteoporosis and fracture. Whether cancer survivors experience a similar post-fracture "care gap" is unclear. Using population-based databases, we assessed whether cancer patients are evaluated and/or treated for osteoporosis after a major fracture.

METHODS

From the Manitoba Cancer Registry, we identified cancer cases (first cancer diagnosis between 1987 and 2013) and cancer-free controls with incident major non-traumatic fractures (from provincial physician billing claims and hospitalization databases). The outcomes were performance of BMD testing (from the BMD Registry), initiation of osteoporosis therapy (from drug dispensation database) or either intervention (BMD testing and/or osteoporosis therapy) in the 12 months post-fracture.

RESULTS

There were 4238 cancer and 16,418 cancer-free individuals who sustained a fracture after the index date (cancer diagnosis) and were followed for at least 1 year post-fracture. Subsequent BMD testing was performed in 11.0% of cancer cases versus 11.5% non-cancer controls (P = 0.43), osteoporosis treatment in 22.9% cancer cases versus 21.8% non-cancer controls (P = 0.15), and either testing or treatment in 28.9% cancer cases versus 28.4% non-cancer controls (P = 0.53). Predictors of BMD testing and/or initiation of therapy were similar for non-cancer and cancer patients. Post-fracture interventions were consistently used more frequently among women, older patients (age 50 years or older), those who sustained fractures in a later calendar period, and (for treatment) after vertebral fracture. Cancer-specific variables (cancer type, years from cancer diagnosis to fracture, specialty of care provider) showed only weak and inconsistent effects.

CONCLUSIONS

A large care gap exists among cancer patients who sustain a fracture, similar to the general population, whereby the evaluation or treatment for osteoporosis is seldom conducted. Care maps may need to be developed for cancer populations to improve post-fracture care.

Canadian Data Platform: Developing an Algorithm Inventory for Health and Social Measures

IntroductionThe SPOR (Strategy for Patient-Oriented Research) Canadian Data Platform aims to facilitate multi-jurisdictional research through a variety of activities, including the development of standardized algorithms for health conditions, health service use, and the determinants of health. An initial step towards standardization was to identify existing health measures that have been validated or assessed for feasibility of implementation in multi-jurisdictional research, document features of these measures, and describe the methods used to validate or assess feasibility. Objectives and ApproachWe constructed an inventory of published algorithms to measure population health, health services use, and the determinants of health. A systematic review of published literature identified algorithms from validation or feasibility studies in two or more Canadian provinces/territories. The search strategy was applied to Medline, Embase, and Scopus. The Algorithms and Harmonized Data Working Group of the Canadian Data Platform identified relevant fields for data extraction, including study type, population characteristics, data source, jurisdictions, and algorithm details. A searchable online resource was created to maintain and share the algorithms. ResultsOf the 2758 articles retrieved, 1998 articles underwent title and abstract review and 60 articles were selected for full review. A total of 8 validation and 26 feasibility studies were assessed; they contributed over 140 algorithms. Chronic physical health conditions, such as diabetes, depression, hypertension and dementia, were most often represented in the algorithms. British Columbia and Manitoba were the jurisdictions most frequently represented in the studies. Methods to facilitate automated searching of the on-line resource are under development. Conclusion / ImplicationsOur inventory of algorithms provides valuable information for researchers interested in conducting multi-jurisdictional studies, and reveals gaps where further algorithm development could be undertaken. This comprehensive collection of existing algorithms will support future studies aimed at improving population health and monitoring health service use in Canada.

Combining population-based administrative health records and electronic medical records for disease surveillance

Background

Administrative health records (AHRs) and electronic medical records (EMRs) are two key sources of population-based data for disease surveillance, but misclassification errors in the data can bias disease estimates. Methods that combine information from error-prone data sources can build on the strengths of AHRs and EMRs. We compared bias and error for four data-combining methods and applied them to estimate hypertension prevalence.

Methods

Our study included rule-based OR and AND methods that identify disease cases from either or both data sources, respectively, rule-based sensitivity-specificity adjusted (RSSA) method that corrects for inaccuracies using a deterministic rule, and probabilistic-based sensitivity-specificity adjusted (PSSA) method that corrects for error using a statistical model. Computer simulation was used to estimate relative bias (RB) and mean square error (MSE) under varying conditions of population disease prevalence, correlation amongst data sources, and amount of misclassification error. AHRs and EMRs for Manitoba, Canada were used to estimate hypertension prevalence using validated case definitions and multiple disease markers.

Results

The OR method had the lowest RB and MSE when population disease prevalence was 10%, and the RSSA method had the lowest RB and MSE when population prevalence increased to 20%. As the correlation between data sources increased, the OR method resulted in the lowest RB and MSE. Estimates of hypertension prevalence for AHRs and EMRs alone were 30.9% (95% CI: 30.6–31.2) and 24.9% (95% CI: 24.6–25.2), respectively. The estimates were 21.4% (95% CI: 21.1–21.7), for the AND method, 34.4% (95% CI: 34.1–34.8) for the OR method, 32.2% (95% CI: 31.8–32.6) for the RSSA method, and ranged from 34.3% (95% CI: 34.1–34.5) to 35.9% (95% CI, 35.7–36.1) for the PSSA method, depending on the statistical model.

Conclusions

The OR and AND methods are influenced by correlation amongst the data sources, while the RSSA method is dependent on the accuracy of prior sensitivity and specificity estimates. The PSSA method performed well when population prevalence was high and average correlations amongst disease markers was low. This study will guide researchers to select a data-combining method that best suits their data characteristics.

Electronic supplementary material

The online version of this article (10.1186/s12911-019-0845-5) contains supplementary material, which is available to authorized users.

Chronic Disease Case Definitions for Electronic Medical Records: A Canadian Validation Study

ABSTRACTObjectivesCanadians are investing heavily in electronic medical records (EMRs) to inform primary care practice improvements. The Canadian Primary Care Sentinel Surveillance Network (CPCSSN) is a national practice-based network that has enrolled more than one million patients to date. Accurate CPCSSN EMR data are essential for unbiased research about chronic disease prevention and management. The study purpose was to test the accuracy of chronic disease case definitions in EMR data from one CPCSSN site. ApproachThis study linked CPCSSN EMR data, hospital records, physician billing claims, prescription drug records, and population registration files for the province of Manitoba. Individuals who had at least one encounter with a CPCSSN practice between 1998 and 2012, were at least 18 years of age, and had a minimum of two years of healthcare coverage before and after the study index date were included. Separate cohorts were defined for the following chronic diseases: chronic obstructive pulmonary disease (COPD), depression, diabetes, hypertension, and osteoarthritis. Validated case definitions based on diagnoses in physician and hospital records and prescription drug data were used estimate sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and kappa of each EMR chronic disease case definition. ResultsMore than 74,000 individuals were included in each cohort, except for COPD which had 51,000. Approximately half of each cohort was comprised of urban residents. The average age ranged from 45.9 years for individuals with depression to 65.3 years for individuals with COPD. Hypertension had the highest prevalence (22.0%) in EMR data followed by depression (14.6%). Estimates of agreement (i.e., kappa) for EMR and administrative data ranged from 0.47 for COPD to 0.58 for diabetes. Sensitivity of the EMR data was lowest for COPD (37.4%; 95% CI 36.0-38.8) and highest for diabetes (57.6%; 95% confidence interval [CI] 56.6-58.6). PPV estimates were lowest for osteoarthritis (66.9%; 95% CI 66.0-67.8) and highest for hypertension (78.3%; 95% CI 77.7-78.9). Specificity estimates were consistently above 90% and NPV estimates were always greater than 80%. Validity estimates for the EMR case definitions were associated with demographic and comorbidity characteristics of the study cohorts. ConclusionsValidity of EMR data, when compared to administrative health data, for ascertaining five different chronic diseases was fair to good; it varied with the disease under investigation. Further research is needed to identify methods for improving the accuracy of chronic disease case definitions in EMR data.