Cohort profile: a large EHR-based cohort with linked pharmacy refill and neighbourhood social determinants of health data to assess heart failure medication adherence

PURPOSE

Clinic-based or community-based interventions can improve adherence to guideline-directed medication therapies (GDMTs) among patients with heart failure (HF). However, opportunities for such interventions are frequently missed, as providers may be unable to recognise risk patterns for medication non-adherence. Machine learning algorithms can help in identifying patients with high likelihood of non-adherence. While a number of multilevel factors influence adherence, prior models predicting non-adherence have been limited by data availability. We have established an electronic health record (EHR)-based cohort with comprehensive data elements from multiple sources to improve on existing models. We linked EHR data with pharmacy refill data for real-time incorporation of prescription fills and with social determinants data to incorporate neighbourhood factors.

PARTICIPANTS

Patients seen at a large health system in New York City (NYC), who were >18 years old with diagnosis of HF or reduced ejection fraction (<40%) since 2017, had at least one clinical encounter between 1 April 2021 and 31 October 2022 and active prescriptions for any of the four GDMTs (beta-blocker, ACEi/angiotensin receptor blocker (ARB)/angiotensin receptor neprilysin inhibitor (ARNI), mineralocorticoid receptor antagonist (MRA) and sodium-glucose cotransporter 2 inhibitor (SGLT2i)) during the study period. Patients with non-geocodable address or outside the continental USA were excluded.

FINDINGS TO DATE

Among 39 963 patients in the cohort, the average age was 73±14 years old, 44% were female and 48% were current/former smokers. The common comorbid conditions were hypertension (77%), cardiac arrhythmias (56%), obesity (33%) and valvular disease (33%). During the study period, 33 606 (84%) patients had an active prescription of beta blocker, 32 626 (82%) had ACEi/ARB/ARNI, 11 611 (29%) MRA and 7472 (19%) SGLT2i. Ninety-nine per cent were from urban metropolitan areas.

FUTURE PLANS

We will use the established cohort to develop a machine learning model to predict medication adherence, and to support ancillary studies assessing associates of adherence. For external validation, we will include data from an additional hospital system in NYC.

Organizational access points and substance use disorder treatment utilization among Black women: a longitudinal cohort study

INTRODUCTION

Health and social service organizations, including the emergency department (ED) and public assistance programs, constitute a social safety net that may serve as an "access point" for substance use treatment utilization. Racialization of substance use disorder (SUD) and gender disparities in access to treatment contribute to differences in health and social service utilization, including substance use treatment for Black women. We therefore explored the role of various access points in facilitating the use of substance use treatment among Black women with substance use and involvement in the criminal justice system.

METHODS

We used data from the Black Women in the Study of Epidemics (B-WISE) project (2008-2011), which recruited Black women who use drugs from community, probation, and prison recruitment settings in Kentucky. B-WISE is a three-wave panel survey collected on a six-month interval. We estimated dynamic panel models to understand whether time-varying use of services influenced women's substance use treatment utilization over 18-months, adjusting for time-invariant characteristics. We stratified the analysis based on where women were recruited (i.e., community, prison, and probation).

RESULTS

The sample included 310 persons and 930 person-waves. For the community and prison samples, the use of an ED in the 6 months prior decreased women's likelihood of subsequent substance use treatment use (Coef: -0.21 (95% CI: -0.40, -0.01); -0.33 (95% CI: -0.60, -0.06), respectively). For the probation sample, receiving support from public assistance (i.e., food stamps, housing, cash assistance) increased the likelihood of subsequent substance use treatment use (0.27 (95% CI: 0.08, 0.46)).

CONCLUSION

Interactions with health and social service organizations predicted Black women's use of substance use treatment services and varied based on their involvement in the criminal justice system. Public assistance venues for Black women on probation may be a point of intervention to increase their access to and use of substance use treatment.

The measurement of symptoms in children with cancer

The purpose of this study was to determine symptom prevalence, characteristics, and distress in children with cancer. The Memorial Symptom Assessment Scale (MSAS) 10-18, a 30-item patient-rated instrument adapted from a previously validated adult version, provided multidimensional information about the symptoms experienced by children with cancer. This instrument was administered to 160 children with cancer aged 10-18 (45 inpatients, 115 outpatients). To confirm the instrument's reliability and validity, additional data about symptoms were collected from both the parents and the medical charts, and retesting was performed on a subgroup of inpatients. Patients could easily complete the scale in a mean of 11 minutes. The analyses supported the reliability and validity of the MSAS 10-18 subscale scores as measures of physical, psychological, and global symptom distress, respectively. Symptom prevalence ranged from 49.7% for lack of energy to 6.3% for problems with urination. The mean (+/- SD) number of symptoms per inpatient was 12.7 +/- 4.9 (range, 4-26), significantly more than the mean 6.5 +/- 5.7 (range, 0-28) symptoms per outpatient. Patients who had recently received chemotherapy had significantly more symptoms than patients who had not received chemotherapy for more than 4 months (11.6 +/- 6.0 vs. 5. 2 +/- 5.1), and those patients with solid tumors had significantly more symptoms than patients with either leukemia, lymphoma, or central nervous system malignancies (9.9 +/- 7.0 vs. 6.8 +/- 5.5 vs. 6.8 +/- 5.0 vs. 8.0 +/- 6.1). The most common symptoms (prevalence > 35%) were lack of energy, pain, drowsiness, nausea, cough, lack of appetite, and psychological symptoms (feeling sad, feeling nervous, worrying, feeling irritable). Of the symptoms with prevalence rates > 35%, those that caused high distress in more than one-third of patients were feeling sad, pain, nausea, lack of appetite, and feeling irritable. Subscale scores demonstrated large variability in symptom distress and could identify subgroups with high distress. The prevalence, characteristics, and distress associated with physical and psychological symptoms could be quantified in older children with cancer. The data confirm a high prevalence of symptoms overall and the existence of subgroups with high distress associated with one or multiple symptoms. Symptom distress is relatively higher among inpatients, children with solid tumors, and children who are undergoing antineoplastic treatment. Systematic symptom assessment may be useful in future epidemiological studies of symptoms and in clinical chemotherapeutic trials. Symptom epidemiology may also provide a focus for future clinical trials related to symptom management in children with cancer.