Pragmatic Trial of Messaging to Providers About Treatment of Hyperlipidemia (PROMPT-LIPID): A Randomized Clinical Trial.

BACKGROUND: Lipid-lowering therapy (LLT) is underutilized for very high-risk atherosclerotic cardiovascular disease. PROMPT-LIPID (PRagmatic Trial of Messaging to Providers about Treatment of HyperLIPIDemia) sought to determine whether electronic health record (EHR) alerts improve 90-day LLT intensification in patients with very high-risk atherosclerotic cardiovascular disease. METHODS: PROMPT-LIPID was a pragmatic trial in which cardiovascular and internal medicine clinicians within Yale New Haven Health (New Haven, CT) were cluster-randomized to receive an EHR alert with individualized LLT recommendations or no alert for outpatients with very high-risk atherosclerotic cardiovascular disease and LDL-C (low-density lipoprotein cholesterol), ≥70 mg/dL. The primary outcome was 90-day LLT intensification (change to high-intensity statin and addition of ezetimibe or PCSK9i [proprotein subtilisin/kexin type 9 inhibitors]). Secondary outcomes included LDL-C level, proportion of patients with LDL-C of <70 or < 55 mg/dL, rate of major adverse cardiovascular events, ED visit incidence, and 6-month mortality. Results were analyzed using logistic and linear regression clustered at the provider level. RESULTS: The no-alert group included 47 clinicians and 1370 patients (median age, 71 years; 50.1% female, median LDL-C, 93 mg/dL); the alert group included 49 clinicians and 1130 patients (median age, 72 years; 47% female, median LDL-C 91, mg/dL). The primary outcome was observed in 14.1% of patients in the alert group as compared with 10.4% in the no-alert group. There were no differences in any secondary outcomes at 6 months. Among 542 patients whose clinicians (n=46) did not dismiss the EHR alert recommendations, LLT intensification was significantly greater (21.2% versus 10.4%, odds ratio, 2.33 [95% CI, 1.48-3.66]). CONCLUSIONS: With a real-time, targeted, individualized EHR alert as compared with usual care, the proportion of patients with atherosclerotic cardiovascular disease with LLT intensification was numerically higher but not statistically significant. Among clinicians who did not dismiss the alert, there was a > 2-fold increase in LLT intensification. EHR alerts, coupled with strategies to reduce clinician dismissal, may help address persistent gaps in LDL-C management. REGISTRATION: URL: https://www.clinicaltrials.gov; Unique identifier: NCT04394715, https://www.clinicaltrials.gov/ct2/show/study/NCT04394715.

Thirty-Day and 90-Day Episode of Care Spending Following Heart Failure Hospitalization Among Medicare Beneficiaries.

BACKGROUND: Despite growing interest in value-based models, utilization patterns and costs for heart failure (HF) admissions are not well understood. We sought to characterize Medicare spending for patients with HF for 30- and 90-day episodes of care (which include an index hospitalization and 30 or 90 days following discharge) and to describe the patterns of post-acute care spending. METHODS: Using Medicare fee-for-service administrative claims data from 2016 to 2018, we performed a retrospective analysis of patients discharged after hospitalization with primary discharge diagnoses of systolic HF, diastolic HF, hypertensive heart disease (HHD) with HF, and HHD with HF and chronic kidney disease. We analyzed coding patterns across these groups over time, median 30- and 90-day payments, and costs allocated to index hospitalization and postacute care. RESULTS: The study included 935 962 patients discharged following hospitalization for HF (systolic HF: 178 603; diastolic HF: 165 156; HHD with HF: 226 929; HHD with HF and chronic kidney disease: 365 274). The proportion of HHD codes increased from 26% of HF hospitalizations in 2016 to 91% in 2018. There was substantial spending on 30-day (median $13 330, interquartile range $9912-$22 489) and 90-day episodes (median $21 658, interquartile range $12 423-$37 630) for HF with significant variation, such that the third quartile of patients incurred costs 3 times the amount of the first quartile. Across all codes, the index hospitalization accounted for ≈70% of 30-day and 45% of 90-day spending. Sixty-one percent of postacute care spending occurred 31 to 90 days following discharge, with readmissions and observation stays (36%) and skilled nursing facilities (27%) comprising the largest categories. CONCLUSIONS: This patient episode-level analysis of contemporary Medicare beneficiaries is the first to examine 90-day spending, which will become an increasingly important pasyment benchmark with the expansion of the Medicare Bundled Payments for Care Improvement Program. Further investigation into the drivers of costs will be essential to provide high-value HF care.