Electronic medical record-assisted design of a cluster-randomized trial to improve diabetes care and outcomes.

BACKGROUND: Electronic medical records (EMRs) have the potential to facilitate the design of large cluster-randomized trials (CRTs). OBJECTIVE: To describe the design of a CRT of clinical decision support to improve diabetes care and outcomes. METHODS: In the Diabetes Improvement Group-Intervention Trial (DIG-IT), we identified and balanced preassignment characteristics of 12,675 diabetic patients cared for by 147 physicians in 24 practices of 2 systems using the same vendor's EMR. EMR-facilitated disease management was system A's experimental intervention; system B interventions involved patient empowerment, with or without disease management. For our sample, we: (1) identified characteristics associated with response to interventions or outcomes; (2) summarized feasible partitions of 10 system A practices (2 groups) and 14 system B practices (3 groups) using intra-cluster correlation coefficients (ICCs) and standardized differences; (3) selected (blinded) partitions to effectively balance the characteristics; and (4) randomly assigned groups of practices to interventions. RESULTS: In System A, 4,306 patients, were assigned to 2 groups of practices; 8,369 patients in system B were assigned to 3 groups of practices. Nearly all baseline outcome variables and covariates were well-balanced, including several not included in the initial design. DIG-IT's balance was superior to alternative partitions based on volume, geography or demographics alone. CONCLUSIONS: EMRs facilitated rigorous CRT design by identifying large numbers of patients with diabetes and enabling fair comparisons through preassignment balancing of practice sites. Our methods can be replicated in other settings and for other conditions, enhancing the power of other translational investigations.

Trends in postdischarge mortality and readmissions: has length of stay declined too far?

BACKGROUND: Length of hospital stay continues to decline, but the effect on postdischarge outcomes is unclear. METHODS: We determined trends in risk-adjusted mortality rates and readmission rates for 83,445 Medicare patients discharged alive after hospitalization for myocardial infarction, heart failure, gastrointestinal hemorrhage, chronic obstructive pulmonary disease, pneumonia, or stroke. Patients were stratified into deciles of observed/expected length of stay to determine whether patients whose length of stay was much shorter than expected had higher risk-adjusted mortality and readmission rates. Analyses were stratified by whether a do-not-resuscitate (DNR) order was written within 2 days of admission (early) or later. RESULTS: From 1991 through 1997, risk-adjusted postdischarge mortality generally remained stable for patients without a DNR order. Postdischarge mortality increased by 21% to 72% for patients with early DNR orders and increased for 2 of 6 diagnoses for patients with late DNR orders. Markedly shorter than expected length of stay was associated with higher than expected risk-adjusted mortality for patients with early DNR orders but not for others (no DNR and late DNR). Risk-adjusted readmission rates remained stable from 1991 through 1997, except for a 15% (95% confidence interval, 3%-30%) increase for patients with congestive heart failure. Short observed/expected length of stay was not associated with higher readmission rates. CONCLUSIONS: The dramatic decline in length of stay from 1991 through 1997 was not associated with worse postdischarge outcomes for patients without DNR orders. However, postdischarge mortality increased among patients with early DNR orders, and some of this trend may be due to patients being discharged more rapidly than previously.

Mortality trends during a program that publicly reported hospital performance.

BACKGROUND: It is unclear whether publicly reporting hospitals' risk-adjusted mortality leads to improvements in outcomes. OBJECTIVES: To examine mortality trends during a period (1991-1997) when the Cleveland Health Quality Choice program was operational. RESEARCH DESIGN: Time series. SUBJECTS: Medicare patients hospitalized with acute myocardial infarction (AMI; n = 10,439), congestive heart failure (CHF; n = 23,505), gastrointestinal hemorrhage (GIH; n = 11,088), chronic obstructive pulmonary disease (COPD; n = 8495), pneumonia (n = 23,719), or stroke (n = 14,293). MEASURES: Risk-adjusted in-hospital mortality, early postdischarge mortality (between discharge and 30 days after admission), and 30-day mortality. RESULTS: Risk-adjusted in-hospital mortality declined significantly for all conditions except stroke and GIH, with absolute declines ranging from -2.1% for COPD to -4.8% for pneumonia. However, the mortality rate in the early postdischarge period rose significantly for all conditions except COPD, with increases ranging from 1.4% for GIH to 3.8% for stroke. As a consequence, the 30-day mortality declined significantly only for CHF (absolute decline 1.4%, 95% CI, -2.5 to -0.1%) and COPD (absolute decline 1.6%, 95% CI, -2.8-0.0%). For stroke, risk-adjusted 30-day mortality actually increased by 4.3% (95% CI, 1.8-7.1%). CONCLUSION: During Cleveland's experiment with hospital report cards, deaths shifted from in hospital to the period immediately after discharge with little or no net reduction in 30-day mortality for most conditions. Hospital profiling remains an unproven strategy for improving outcomes of care for medical conditions. Using in-hospital mortality rates to monitor trends in outcomes for hospitalized patients may lead to spurious conclusions.