Targeted learning in observational studies with multi-valued treatments: An evaluation of antipsychotic drug treatment safety

We investigate estimation of causal effects of multiple competing (multi-valued) treatments in the absence of randomization. Our work is motivated by an intention-to-treat study of the relative cardiometabolic risk of assignment to one of six commonly prescribed antipsychotic drugs in a cohort of nearly 39 000 adults with serious mental illnesses. Doubly-robust estimators, such as targeted minimum loss-based estimation (TMLE), require correct specification of either the treatment model or outcome model to ensure consistent estimation; however, common TMLE implementations estimate treatment probabilities using multiple binomial regressions rather than multinomial regression. We implement a TMLE estimator that uses multinomial treatment assignment and ensemble machine learning to estimate average treatment effects. Our multinomial implementation improves coverage, but does not necessarily reduce bias, relative to the binomial implementation in simulation experiments with varying treatment propensity overlap and event rates. Evaluating the causal effects of the antipsychotics on 3-year diabetes risk or death, we find a safety benefit of moving from a second-generation drug considered among the safest of the second-generation drugs to an infrequently prescribed first-generation drug known for having low cardiometabolic risk.

Healthcare Access for a Diverse Population with Schizophrenia Following the Onset of the COVID-19 Pandemic

COVID-19 has had a disproportionate impact on the most disadvantaged members of society, including minorities and those with disabling chronic illnesses such as schizophrenia. We examined the pandemic's impacts among New York State's Medicaid beneficiaries with schizophrenia in the immediate post-pandemic surge period, with a focus on equity of access to critical healthcare. We compared changes in utilization of key behavioral health outpatient services and inpatient services for life-threatening conditions between the pre-pandemic and surge periods for White and non-White beneficiaries. We found racial and ethnic differences across all outcomes, with most differences stable over time. The exception was pneumonia admissions-while no differences existed in the pre-pandemic period, Black and Latinx beneficiaries were less likely than Whites to be hospitalized in the surge period despite minorities' heavier COVID-19 disease burden. The emergence of racial and ethnic differences in access to scarce life-preserving healthcare may hold lessons for future crises.

Antipsychotics and the risk of diabetes and death among adults with serious mental illnesses

BACKGROUND

Individuals with schizophrenia exposed to second-generation antipsychotics (SGA) have an increased risk for diabetes, with aripiprazole purportedly a safer drug. Less is known about the drugs' mortality risk or whether serious mental illness (SMI) diagnosis or race/ethnicity modify these effects.

METHODS

Authors created a retrospective cohort of non-elderly adults with SMI initiating monotherapy with an SGA (olanzapine, quetiapine, risperidone, and ziprasidone, aripiprazole) or haloperidol during 2008-2013. Three-year diabetes incidence or all-cause death risk differences were estimated between each drug and aripiprazole, the comparator, as well as effects within SMI diagnosis and race/ethnicity. Sensitivity analyses evaluated potential confounding by indication.

RESULTS

38 762 adults, 65% White and 55% with schizophrenia, initiated monotherapy, with haloperidol least (6%) and quetiapine most (26·5%) frequent. Three-year mortality was 5% and diabetes incidence 9.3%. Compared with aripiprazole, haloperidol and olanzapine reduced diabetes risk by 1.9 (95% CI 1.2-2.6) percentage points, or a 18.6 percentage point reduction relative to aripiprazole users' unadjusted risk (10.2%), with risperidone having a smaller advantage. Relative to aripiprazole users' unadjusted risk (3.4%), all antipsychotics increased mortality risk by 1.1-2.2 percentage points, representing 32.4-64.7 percentage point increases. Findings within diagnosis and race/ethnicity were generally consistent with overall findings. Only quetiapine's higher mortality risk held in sensitivity analyses.

CONCLUSIONS

Haloperidol's, olanzapine's, and risperidone's lower diabetes risks relative to aripiprazole were not robust in sensitivity analyses but quetiapine's higher mortality risk proved robust. Findings expand the evidence on antipsychotics' risks, suggesting a need for caution in the use of quetiapine among individuals with SMI.

Statistical Approaches for Quantifying the Quality of Neurosurgical Care

BACKGROUND

Quantifying quality of health care can provide valuable information to patients, providers, and policy makers. However, the observational nature of measuring quality complicates assessments.

METHODS

We describe a conceptual model for defining quality and its implications about the data collected, how to make inferences about quality, and the assumptions required to provide statistically valid estimates. Twenty-one binary or polytomous quality measures collected from 101,051 adult Medicaid beneficiaries aged 18-64 years with schizophrenia from 5 U.S. states show methodology. A categorical principal components analysis establishes dimensionality of quality, and item response theory models characterize the relationship between each quality measure and a unidimensional quality construct. Latent regression models estimate racial/ethnic and geographic quality disparities.

RESULTS

More than 90% of beneficiaries filled at least 1 antipsychotic prescription and 19% were hospitalized for schizophrenia during a 12-month observational period in our multistate cohort with approximately 2/3 nonwhite beneficiaries. Four quality constructs emerged: inpatient, emergency room, pharmacologic/ambulatory, and ambulatory only. Using a 2-parameter logistic model, pharmacologic/ambulatory care quality varied from -2.35 to 1.26 (higher = better quality). Black and Latinx beneficiaries had lower pharmacologic/ambulatory quality compared with whites. Race/ethnicity modified the association of state and pharmacologic/ambulatory care quality in latent regression modeling. Average quality ranged from -0.28 (95% confidence interval, -2.15 to 1.04) for blacks in New Jersey to 0.46 [95% confidence interval, -0.89 to 1.40] for whites in Michigan.

CONCLUSIONS

By combining multiple quality measures using item response theory models, a composite measure can be estimated that has more statistical power to detect differences among subjects than the observed mean per subject.

Mortality Prediction Following Cardiac Surgery in Children - An STS Congenital Heart Surgery Database Analysis

BACKGROUND

The Society of Thoracic Surgeons' Congenital Heart Surgery Database (STS CHSD) provides risk-adjusted operative mortality rates to approximately 120 North American congenital heart centers. Optimal case-mix adjustment methods for operative mortality risk prediction in this population remain unclear.

METHODS

A panel created diagnosis-procedure (D-P) combinations of encounters in the CHSD. Models for operative mortality using the new D-P categories, procedure-specific risk factors, and syndromes/abnormalities included in the CHSD were estimated using Bayesian additive regression trees (BART) and lasso models. Performance of the new models was compared to the current STS-CHSD risk model.

RESULTS

Of 98,825 operative encounters (69,063 training; 29,762 testing), 2,818 (2.85%) STS-defined operative mortalities were observed. Differences in sensitivity, specificity, true and false positive predicted values were negligible across models. Calibration for mortality predictions at the higher end of risk from the lasso and BART models was better than predictions from the STS-CHSD model, likely due to new models' inclusion of diagnosis-palliative procedure variables affecting < 1% of patients overall, but accounting for 27% of mortalities. Model discrimination varied across models for high-risk procedures, hospital volume, and hospitals.

CONCLUSIONS

Overall performance of the new models did not differ meaningfully from the STS-CHSD risk model. Addition of procedure-specific risk factors and allowing diagnosis to modify predicted risk for palliative operations may augment model performance for very high-risk surgeries. Given the importance of risk adjustment in estimating hospital quality, a comparative assessment of surgical program quality evaluations using the different models is warranted.

Factors Associated With Off-Label Utilization of Second-Generation Antipsychotics Among Publicly Insured Adults

OBJECTIVE

Off-label utilization of second-generation antipsychotic medications may expose patients to significant risks. The authors examined the prevalence, temporal trends, and factors associated with off-label utilization of second-generation antipsychotics among publicly insured adults.

METHODS

A retrospective repeated panel was used to examine monthly off-label utilization of second-generation antipsychotics among fee-for-service Medicare, Medicaid, and dually eligible White, Black, and Latino adult beneficiaries filling prescriptions for second-generation antipsychotics in California, Georgia, Mississippi, and Oklahoma from July 2008 through June 2013.

RESULTS

Among 301,367 users of second-generation antipsychotics, between 36.5% and 41.9% had utilization that was always off-label. Payer did not modify effects of race-ethnicity on off-label utilization. Compared with Whites, Blacks had lower monthly odds of off-label utilization in all four states, and Latinos had lower odds of utilization in California and Georgia. Payer was associated with off-label utilization in California, Mississippi, and Oklahoma. California Medicaid beneficiaries were 1.12 (95% confidence interval=1.10-1.13) times as likely as dually eligible beneficiaries to have off-label utilization. Off-label utilization increased relative to the baseline year in all states, but a downward trend followed in three states.

CONCLUSIONS

Off-label utilization of second-generation antipsychotics was prevalent despite the drugs' cardiometabolic risks and little evidence of their effectiveness. The lower likelihood of off-label utilization among patients from racial-ethnic minority groups might stem from prescribers' efforts to minimize risks, given a higher baseline risk for these groups, or from disparities-associated factors. Variation among payers suggests that payer policies can affect off-label utilization.

Significance and Factors Associated with Antipsychotic Polypharmacy Utilization Among Publicly Insured US Adults

Antipsychotic polypharmacy (APP) lacks evidence of effectiveness in the care of schizophrenia or other disorders for which antipsychotic drugs are indicated, also exposing patients to more risks. Authors assessed APP prevalence and APP association with beneficiary race/ethnicity and payer among publicly-insured adults regardless of diagnosis. Retrospective repeated panel study of fee-for-service (FFS) Medicare, Medicaid, and dually-eligible white, black, and Latino adults residing in California, Georgia, Iowa, Mississippi, Oklahoma, South Dakota, or West Virginia, filling antipsychotic prescriptions between July 2008 and June 2013. Primary outcome was any monthly APP utilization. Across states and payers, 11% to 21% of 397,533 antipsychotic users and 12% to 19% of 9,396,741 person-months had some APP utilization. Less than 50% of person-months had a schizophrenia diagnosis and up to 19% had no diagnosed mental illness. Payer modified race/ethnicity effects on APP utilization only in CA; however, the odds of APP utilization remained lower for minorities than for whites. Elsewhere, the odds varied by race/ethnicity only in OK, with Latinos having lower odds than whites (odds ratio 0.76; 95% confidence interval 0.60-0.96). The odds of APP utilization varied by payer in several study states, with odds generally higher for Dual eligibles, although the differences were generally small; the odds also varied by year (lower at study end). APP was frequently utilized but mostly declined over time. APP utilization patterns varied across states, with no consistent association with race/ethnicity and small payer effects. Greater use of APP-reducing strategies are needed, particularly among non-schizophrenia populations.

All-Cause Mortality and Progression to End-Stage Kidney Disease Following Percutaneous Revascularization or Surgical Coronary Revascularization in Patients with CKD

Introduction

Relative impacts of coronary artery bypass grafting (CABG) and percutaneous coronary intervention (PCI) on mortality and end-stage kidney disease (ESKD) in chronic kidney disease (CKD) are uncertain.

Methods

Data from Massachusetts residents with CKD undergoing CABG or PCI from 2003 to 2012 were linked to the United States Renal Data System. Associations with death, ESKD, and combined death and ESKD were analyzed in propensity score−matched multivariable survival models.

Results

We identified 6805 CABG and 17,494 PCI patients. Among 3775 matched-pairs, multi-vessel disease was present in 97%, and stage 4 CKD was present in 11.9% of CABG and 12.2% of PCI patients. One-year mortality (CABG 7.7%, PCI 11.0%) was more frequent than ESKD (CABG 1.4%, PCI 1.7%). Overall survival was improved and ESKD risk decreased with CABG compared to PCI, but effects differed in the presence of left main disease and prior myocardial infarction (MI). Survival was worse following PCI than following CABG among patients with left main disease and without MI (hazard ratio = 3.7, 95% confidence interval = 1.3−10.5). ESKD risk was higher with PCI for individuals with left main disease and prior infarction (hazard ratio = 8.1, 95% confidence interval = 1.7−39.2).

Conclusion

Risks following CABG and PCI were modified by left main disease and prior MI. In individuals with CKD, survival was greater after CABG than after PCI in patients with left main disease but without MI, whereas ESKD risk was lower with CABG in those with left main and MI. Absolute risks of ESKD were markedly lower than for mortality, suggesting prioritizing mortality over ESKD in clinical decision making.

Assessment of Second-Generation Diabetes Medication Initiation Among Medicare Enrollees From 2007 to 2015

IMPORTANCE

Little is known about how new and expensive drugs diffuse into practice affects health care costs.

OBJECTIVE

To describe the variation in second-generation diabetes drug use among Medicare enrollees between 2007 and 2015.

DESIGN, SETTING, AND PARTICIPANTS

This population-based, cross-sectional study included data from 100% of Medicare Parts A, B, and D enrollees who first received diabetes drug therapy from January 1, 2007, to December 31, 2015. Patients with type 1 diabetes were excluded. Data were analyzed beginning in the spring of 2018, and revisions were completed in 2019.

EXPOSURES

For each patient, the initial diabetes drug choice was determined; drugs were classified as first generation (ie, approved before 2000) or second generation (ie, approved after 2000, including dipeptidyl peptidase 4 [DPP-4] inhibitors, glucagon-like peptide-1 [GLP-1] receptor agonists, and sodium-glucose cotransporter-2 [SGLT-2] inhibitors).

MAIN OUTCOMES AND MEASURES

The primary outcome was the between-practice variation in use of second-generation diabetes drugs between 2007 and 2015. Practices with use rates of second-generation diabetes drugs more than 1 SD above the mean were considered high prescribing, while those with use rates more than 1 SD below the mean were considered low prescribing.

RESULTS

Among 1 182 233 patients who initiated diabetes drug therapy at 42 977 practices between 2007 and 2015, 1 104 718 (93.4%) were prescribed a first-generation drug (mean [SD] age, 75.4 [6.7] years; 627 134 [56.8%] women) and 77 515 (6.6%) were prescribed a second-generation drug (mean [SD] age, 76.5 [7.2] years; 44 697 [57.7%] women). By December 2015, 22 457 practices (52.2%) had used DPP-4 inhibitors once, compared with 3593 practices (8.4%) that had used a GLP-1 receptor agonist once. Furthermore, 17 452 practices (40.6%) were using DPP-4 inhibitors in 10% of eligible patients, while 1286 practices (3.0%) were using GLP-1 receptor agonists in 10% of eligible patients, and SGLT-2 inhibitors, available after March 2013, were used at least once by 1716 practices (4.0%) and used in 10% of eligible patients by 872 practices (2.0%) by December 2015. According to Poisson random-effect regression models, beneficiaries in high-prescribing practices were more than 3-fold more likely to receive DPP-4 inhibitors (relative risk, 3.55 [95% CI, 3.42-3.68]), 24-fold more likely to receive GLP-1 receptor agonists (relative risk, 24.06 [95% CI, 14.14-40.94]) and 60-fold more likely to receive SGLT-2 inhibitors (relative risk, 60.41 [95% CI, 15.99-228.22]) compared with beneficiaries in low-prescribing practices.

CONCLUSIONS AND RELEVANCE

These findings suggest that there was substantial between-practice variation in the use of second-generation diabetes drugs between 2007 and 2015, with a concentration of use among a few prescribers and practices responsible for much of the early diffusion.

Predicting Participation in Psychiatric Randomized Controlled Trials: Insights From the STEP-BD

OBJECTIVE

Differences between patients who do and do not participate in randomized controlled trials (RCTs) could diminish the generalizability of results. This study examined whether RCT participants differ from non-RCT participants who are recruited from the same patient and provider population.

METHODS

The Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD) was an observational study in which participants also could enroll in an RCT during exacerbations of acute depression. The odds that a patient was enrolled in the STEP-BD acute depression RCTs (pharmacotherapy or psychotherapy) were estimated by fitting logistic regression models to STEP-BD participants with acute bipolar depression (total N=2,222; RCT, N=413; observational arm, N=1,809). Predictor variables included demographic characteristics, clinical information (including severity scales and comorbidities), and study site. The extent to which site determined RCT participation was estimated by using the area under the receiver operating characteristic curve (AUC).

RESULTS

RCT participation was associated with having no insurance (odds ratio [OR]=1.58, 95% confidence interval [CI]=1.16-2.15), a Clinical Global Impression score indicating greater severity (severe versus mild: OR=1.52, CI=1.08-2.15), and site (predicted probability range 8%-31%). Site was the most significant predictor of RCT enrollment (model excluding site, AUC=.61, CI=.58-.64; full model, AUC=.70, CI=.67-.73).

CONCLUSIONS

STEP-BD RCT participants differed from those in the observational arm in few clinical or demographic characteristics. Site was the strongest predictor of RCT participation. Future study is needed to understand site characteristics associated with RCT participation and whether these characteristics are associated with patient outcomes and to test these findings in usual-care settings.

Causes of short-term readmission after percutaneous coronary intervention

BACKGROUND

Rehospitalization within 30 days after an admission for percutaneous coronary intervention (PCI) is common, costly, and a future target for Medicare penalties. Causes of readmission after PCI are largely unknown.

METHODS AND RESULTS

To illuminate the causes of PCI readmissions, patients with PCI readmitted within 30 days of discharge between 2007 and 2011 at 2 hospitals were identified, and their medical records were reviewed. Of 9288 PCIs, 9081 (97.8%) were alive at the end of the index hospitalization. Of these, 893 patients (9.8%) were readmitted within 30 days of discharge and included in the analysis. Among readmitted patients, 341 patients (38.1%) were readmitted for evaluation of recurrent chest pain or other symptoms concerning for angina, whereas 59 patients (6.6%) were readmitted for staged PCI without new symptoms. Complications of PCI accounted for 60 readmissions (6.7%). For cases in which chest pain or other symptoms concerning for angina prompted the readmission, 21 patients (6.2%) met criteria for myocardial infarction, and repeat PCI was performed in 54 patients (15.8%). The majority of chest pain patients (288; 84.4%) underwent ≥1 diagnostic imaging test, most commonly coronary angiography, and only 9 (2.6%) underwent target lesion revascularization.

CONCLUSIONS

After PCI, readmissions within 30 days were seldom related to PCI complications but often for recurrent chest pain. Readmissions with recurrent chest pain infrequently met criteria for myocardial infarction but were associated with high rates of diagnostic testing.

A prediction model to identify patients at high risk for 30-day readmission after percutaneous coronary intervention

BACKGROUND

The Affordable Care Act creates financial incentives for hospitals to minimize readmissions shortly after discharge for several conditions, with percutaneous coronary intervention (PCI) to be a target in 2015. We aimed to develop and validate prediction models to assist clinicians and hospitals in identifying patients at highest risk for 30-day readmission after PCI.

METHODS AND RESULTS

We identified all readmissions within 30 days of discharge after PCI in nonfederal hospitals in Massachusetts between October 1, 2005, and September 30, 2008. Within a two-thirds random sample (Developmental cohort), we developed 2 parsimonious multivariable models to predict all-cause 30-day readmission, the first incorporating only variables known before cardiac catheterization (pre-PCI model), and the second incorporating variables known at discharge (Discharge model). Models were validated within the remaining one-third sample (Validation cohort), and model discrimination and calibration were assessed. Of 36,060 PCI patients surviving to discharge, 3760 (10.4%) patients were readmitted within 30 days. Significant pre-PCI predictors of readmission included age, female sex, Medicare or State insurance, congestive heart failure, and chronic kidney disease. Post-PCI predictors of readmission included lack of β-blocker prescription at discharge, post-PCI vascular or bleeding complications, and extended length of stay. Discrimination of the pre-PCI model (C-statistic=0.68) was modestly improved by the addition of post-PCI variables in the Discharge model (C-statistic=0.69; integrated discrimination improvement, 0.009; P<0.001).

CONCLUSIONS

These prediction models can be used to identify patients at high risk for readmission after PCI and to target high-risk patients for interventions to prevent readmission.

Abstract 321: A Validated Model to Identify Patients at High Risk for 30-Day Readmission After Percutaneous Coronary Intervention

Background: The Affordable Care Act (ACA) creates incentives within Medicare for hospitals that minimize readmissions shortly after discharge. Percutaneous coronary intervention (PCI) has among the highest rates of 30-day all-cause readmission. We developed and validated a prediction model to assist clinicians in identifying patients at high risk for 30-day readmission after discharge for PCI.

Methods: We included all PCI admissions in non-federal hospitals in Massachusetts between October 1, 2005 and September 30, 2008. Readmissions within 30-days of discharge were identified via linkage with Massachusetts inpatient claims files. Within a 2/3 random sample, we developed 2 separate multivariable models to predict all-cause 30-day readmission, one incorporating only variables known prior to cardiac catheterization (Pre-PCI model), and a second incorporating variables known at discharge, including PCI-related complications and discharge disposition (Discharge model). In order to facilitate clinical use of the model via a web-based application, less influential variables were eliminated via stepwise selection, while retaining 95% of the predicted variability in the complete models. Models were validated within the remaining 1/3 sample, and model discrimination and calibration were assessed. Readmissions for staged PCIs were not considered as a readmission.

Results: Of 36060 PCI patients surviving to discharge, 3760 (10.4%) were readmitted within 30 days. In the pre-PCI model, significant independent predictors of readmission included history of heart failure as well as heart failure status at time of PCI, gender, chronic lung disease, worse renal function, insurance status, admission status, previous CABG, peripheral vascular disease, presence of cardiogenic shock, and age. Additional predictors of readmission in the discharge model included length of stay, bleeding or vascular complications, use of drug-eluting stents, previous PCI, diabetes status, race, discharge location, and beta blocker being prescribed at discharge (Figure 1). Model discrimination was moderate for the pre-PCI model (C statistic = 0.67) and not substantially improved by the addition of post-PCI variables (C statistic = 0.69). Both models were well-calibrated within the validation dataset (Hosmer-Lemeshow goodness of fit P = NS for both).

Conclusions: These validated models, developed in a large and broadly generalizable population, can be used to identify patients at high risk for readmission after PCI. Such a model could be used to target high risk patients for interventions to prevent readmission.

Abstract 10: Hospital-Quality Following Percutaneous Coronary Intervention In Massachusetts: Experience From 2003-2008

Background: Little information is available on hospital performance following percutaneous coronary intervention (PCI) using all-comer populations. Using a fully Bayesian approach, Massachusetts has publicly reported all-cause in-hospital mortality (IHM) following PCI procedures performed since 04/01/2003 in all non-federal hospitals.

Methods: Trained data managers prospectively collected information using the ACC-NCDR instrument supplemented by Mass-DAC specific elements. Operators and data managers adjudicated selected elements via chart review. Bayesian hierarchical regression models of IHM were estimated annually for two cohorts: (1) cardiogenic shock or ST-elevation myocardial infarction (STEMI); and (2) no shock and no STEMI. Posterior risk-standardized IHM rates were estimated. The odds of death for a patient treated at a hospital 1 SD below average quality relative to treatment at a hospital 1 SD above average quality quantified variation.

Results: Presence of shock; emergent or salvage status; and compassionate use were changed after adjudication for more than one-quarter of adjudicated cases. No Shock and No STEMI Cohort: Between 04/03-09/08, the number of annual PCI admissions ranged from 11121-14504; number of hospitals from 14-22; and median post-PCI length of stay (LOS) remained unchanged (2 days). The crude IHM declined from 0.76% (81 of 10689) in 2003 to 0.50% (56 in 11275)in 2007. Model discrimination using 9 risk factors was high (ROC = 0.86). The odds of death at lower versus higher quality hospital ranged between 1.38 and 2.14. Shock or STEMI Cohort: the number of PCI admissions ranged from 2606-2800; number of hospitals from 18-24; and median post-PCI LOS remained unchanged (4 days). IHM declined from 6.86% (135 in 1968) in 2003 to 4.78% (130 in 2721)in 2008. Model discrimination was excellent using 7 covariates (ROC = 0.89). The odds of death at lower versus higher quality hospital ranged from 1.49 to 6.35.

Conclusions: Key clinical factors, including shock, status of PCI, and compassionate use require adjudication. A small number of risk factors have excellent discrimination of in-hospital survivors at the patient level. Estimates of between-hospital variation can be used to quantify performance at the hospital level.

Impact of Independent Data Adjudication on Hospital-Specific Estimates of Risk-Adjusted Mortality Following Percutaneous Coronary Interventions in Massachusetts

Background—

As part of state-mandated public reporting of outcomes after percutaneous coronary interventions (PCIs) in Massachusetts, procedural and clinical data were prospectively collected. Variables associated with higher mortality were audited to ensure accuracy of coding. We examined the impact of adjudication on identifying hospitals with possible deficiencies in the quality of PCI care.

Methods and Results—

From October 2005 to September 2006, 15 721 admissions for PCI occurred in 21 hospitals. Of the 864 high-risk variables from 822 patients audited by committee, 201 were changed, with reassignment to lower acuities in 97 (30%) of the 321 shock cases, 24 (43%) of the 56 salvage cases, and 73 (15%) of the 478 emergent cases. Logistic regression models were used to predict patient-specific in-hospital mortality. Of 241 (1.5%) patients who died after PCI, 30 (12.4%) had a lower predicted mortality with adjudicated than with unadjudicated data. Model accuracy was excellent with either adjudicated or unadjudicated data. Hospital-specific risk-standardized mortality rates were estimated using both adjudicated and unadjudicated data through hierarchical logistic regression. Although adjudication reduced between-hospital variation by one third, risk-standardized mortality rates were similar using unadjudicated and adjudicated data. None of the hospitals were identified as statistical outliers. However, cross-validated posterior-predicted P values calculated with adjudicated data increased the number of borderline hospital outliers compared with unadjudicated data.

Conclusions—

Independent adjudication of site-reported high-risk features may increase the ability to identify hospitals with higher risk-adjusted mortality after PCI despite having little impact on the accuracy of risk prediction for the entire population.

Long-term clinical outcomes after drug-eluting and bare-metal stenting in Massachusetts

BACKGROUND

Drug-eluting stents (DES) reduce the need for repeat revascularization, but their long-term safety relative to that of bare-metal stents (BMS) in general use remains uncertain. We sought to compare the clinical outcome of patients treated with DES with that of BMS.

METHODS AND RESULTS

All adults undergoing percutaneous coronary intervention with stenting between April 1, 2003, and September 30, 2004, at non-US government hospitals in Massachusetts were identified from a mandatory state database. Patients were classified from the index admission according to stent types used. Clinical and procedural risk factors were collected prospectively. Risk-adjusted mortality, myocardial infarction, and revascularization rate differences (DES-BMS) were estimated through propensity score matching without replacement. A total of 11 556 patients were treated with DES, and 6237 were treated with BMS, with unadjusted 2-year mortality rates of 7.0% and 12.6%, respectively (P<0.0001). In 5549 DES patients matched to 5549 BMS patients, 2-year risk-adjusted mortality rates were 9.8% and 12.0%, respectively (P=0.0002), whereas the respective rates for myocardial infarction and target-vessel revascularization were 8.3% versus 10.3% (P=0.0005) and 11.0% versus 16.8% (P<0.0001).

CONCLUSIONS

DES treatment was associated with lower rates of mortality, myocardial infarction, and target-vessel revascularization than BMS treatment in similar patients in a matched population-based study. Comprehensive follow-up in this inclusive population is warranted to identify whether similar safety and efficacy remain beyond 2 years.

Drug-eluting or bare-metal stents for acute myocardial infarction

BACKGROUND

Studies comparing percutaneous coronary intervention (PCI) with drug-eluting and bare-metal coronary stents in acute myocardial infarction have been limited in size and duration.

METHODS

We identified all adults undergoing PCI with stenting for acute myocardial infarction between April 1, 2003, and September 30, 2004, at any acute care, nonfederal hospital in Massachusetts with the use of a state-mandated database of PCI procedures. We performed propensity-score matching on three groups of patients: all patients with acute myocardial infarction, all those with acute myocardial infarction with ST-segment elevation, and all those with acute myocardial infarction without ST-segment elevation. Propensity-score analyses were based on clinical, procedural, hospital, and insurance information collected at the time of the index procedure. Differences in the risk of death between patients receiving drug-eluting stents and those receiving bare-metal stents were determined from vital-statistics records.

RESULTS

A total of 7217 patients were treated for acute myocardial infarction (4016 with drug-eluting stents and 3201 with bare-metal stents). According to analysis of matched pairs, the 2-year, risk-adjusted mortality rates were lower for drug-eluting stents than for bare-metal stents among all patients with myocardial infarction (10.7% vs. 12.8%, P=0.02), among patients with myocardial infarction with ST-segment elevation (8.5% vs. 11.6%, P=0.008), and among patients with myocardial infarction without ST-segment elevation (12.8% vs. 15.6%, P=0.04). The 2-year, risk-adjusted rates of recurrent myocardial infarction were reduced in patients with myocardial infarction without ST-segment elevation who were treated with drug-eluting stents, and repeat revascularization rates were significantly reduced with the use of drug-eluting stents as compared with bare-metal stents in all groups.

CONCLUSIONS

In patients presenting with acute myocardial infarction, treatment with drug-eluting stents is associated with decreased 2-year mortality rates and a reduction in the need for repeat revascularization procedures as compared with treatment with bare-metal stents.

Nurse Staffing In Hospitals: Is There A Business Case For Quality?

We construct national estimates of the cost of increasing hospital nurse staffing and associated reductions in days, deaths, and adverse outcomes. Raising the proportion of nursing hours provided by registered nurses (RNs) without increasing total nursing hours is associated with a net reduction in costs. Increasing nursing hours, with or without increasing the proportion of hours provided by RNs, reduces days, adverse outcomes, and patient deaths, but with a net increase in hospital costs of 1.5 percent or less at the staffing levels modeled. Whether or not staffing should be increased depends on the value patients and payers assign to avoided deaths and complications.

Measuring hospital quality: can medicare data substitute for all-payer data?

OBJECTIVES

To assess whether adverse outcomes in Medicare patients can be used as a surrogate for measures from all patients in quality-of-care research using administrative datasets.

DATA SOURCES

Patient discharge abstracts from state data systems for 799 hospitals in 11 states. National MedPAR discharge data for Medicare patients from 3,357 hospitals. State hospital staffing surveys or financial reports. American Hospital Association Annual Survey.

STUDY DESIGN

We calculate rates for 10 adverse patient outcomes, examine the correlation between all-patient and Medicare rates, and conduct negative binomial regressions of counts of adverse outcomes on expected counts, hospital nurse staffing, and other variables to compare results using all-patient and Medicare patient data.

DATA COLLECTION/EXTRACTION

Coding rules were established for eight adverse outcomes applicable to medical and surgical patients plus two outcomes applicable only to surgical patients. The presence of these outcomes was coded for 3 samples: all patients in the 11-state sample, Medicare patients in the 11-state sample, and Medicare patients in the national Medicare MedPAR sample. Logistic regression models were used to construct estimates of expected counts of the outcomes for each hospital. Variables for teaching, metropolitan status, and bed size were obtained from the AHA Annual Survey.

PRINCIPAL FINDINGS

For medical patients, Medicare rates were consistently higher than all-patient rates, but the two were highly correlated. Results from regression analysis were consistent across the 11-state all-patient, 11-state Medicare, and national Medicare samples. For surgery patients, Medicare rates were generally higher than all-patient rates, but correlations of Medicare and all-patient rates were lower, and regression results less consistent.

CONCLUSIONS

Analyses of quality of care for medical patients using Medicare-only and all-patient data are likely to have similar findings. Measures applied to surgery patients must be used with more caution, as those tested only in Medicare patients may not provide results comparable to those from all-patient samples or across different samples of Medicare patients.

Evaluating the Role of Patient Sample Definitions for Quality Indicators Sensitive to Nurse Staffing Patterns

BACKGROUND

Administrative data are an attractive data source for the construction of quality indicators to assess and monitor quality of nursing care in hospitals. Current approaches to constructing measures from discharge abstracts apply substantial restrictions to exclude patients at high risk or with preexisting conditions. This study evaluates whether broader sample definitions combined with risk adjustment would allow for larger samples and increase analytic power.

METHODS

Eight indicators were constructed from discharge abstracts of major surgical and medical patients from 799 hospitals in 11 states using existing definitions: pneumonia, urinary tract infection, decubitus ulcers, central nervous system complications, shock, sepsis, pulmonary failure, and upper gastrointestinal bleeding. We tested the effect of broadening the samples in 4 ways: comparing indicator rates in the broader and restrictive samples; assessing correlations of hospital ranks in the broader and restrictive samples; performing clinical reviews of cases in the added samples; and using different samples in regressions of indicators on nurse staffing variables, adjusting for patient risk.

RESULTS

Indicator rates in the broader samples tended to be higher but did not change hospital rankings significantly. Clinical review suggested that many sample restrictions could be dropped. Using indicators based on broader definitions, coefficients on staffing variables increased in magnitude.

CONCLUSION

Less restrictive sample definitions were shown to be feasible and increased the sensitivity of the indicators and thus the power of the analysis. Particularly in surgical patients, the samples could be broadened, although more conservative definitions appeared appropriate for medical patients.

Nurse-staffing levels and the quality of care in hospitals

BACKGROUND

It is uncertain whether lower levels of staffing by nurses at hospitals are associated with an increased risk that patients will have complications or die.

METHODS

We used administrative data from 1997 for 799 hospitals in 11 states (covering 5,075,969 discharges of medical patients and 1,104,659 discharges of surgical patients) to examine the relation between the amount of care provided by nurses at the hospital and patients' outcomes. We conducted regression analyses in which we controlled for patients' risk of adverse outcomes, differences in the nursing care needed for each hospital's patients, and other variables.

RESULTS

The mean number of hours of nursing care per patient-day was 11.4, of which 7.8 hours were provided by registered nurses, 1.2 hours by licensed practical nurses, and 2.4 hours by nurses' aides. Among medical patients, a higher proportion of hours of care per day provided by registered nurses and a greater absolute number of hours of care per day provided by registered nurses were associated with a shorter length of stay (P=0.01 and P<0.001, respectively) and lower rates of both urinary tract infections (P<0.001 and P=0.003, respectively) and upper gastrointestinal bleeding (P=0.03 and P=0.007, respectively). A higher proportion of hours of care provided by registered nurses was also associated with lower rates of pneumonia (P=0.001), shock or cardiac arrest (P=0.007), and "failure to rescue," which was defined as death from pneumonia, shock or cardiac arrest, upper gastrointestinal bleeding, sepsis, or deep venous thrombosis (P=0.05). Among surgical patients, a higher proportion of care provided by registered nurses was associated with lower rates of urinary tract infections (P=0.04), and a greater number of hours of care per day provided by registered nurses was associated with lower rates of "failure to rescue" (P=0.008). We found no associations between increased levels of staffing by registered nurses and the rate of in-hospital death or between increased staffing by licensed practical nurses or nurses' aides and the rate of adverse outcomes.

CONCLUSIONS

A higher proportion of hours of nursing care provided by registered nurses and a greater number of hours of care by registered nurses per day are associated with better care for hospitalized patients.