Oncology practices often serve as the “medical home” for patients but may not have systems to support all aspects of patient-centered care. In this project, we pilot tested a new set of standards that call for accessible, continuous, coordinated, and team-based care that focuses on the whole person.
(Aim 1) Determine whether the patient-centered oncology standards improve patient experiences and quality and reduces emergency department (ED) and hospital utilization. (Aim 2) Examine how adoption of the standards varies across a variety of practices and which factors affect adoption using qualitative methods.
Five medical oncology practices in Southeastern Pennsylvania implemented the standards from 2014-2016. Implementation support included training webinars and technical assistance. A retrospective, pre-post study design with a concurrent nonrandomized control group of 18 oncology practices addressed aim 1. We used insurance claims to calculate rates of per member per month (PMPM) all-cause hospitalizations, ED use, primary care visits and visits to any specialist type (n = 28 826 eligible patients during baseline and 30 843 during follow-up) and to identify patients who received drug therapy for a cancer care experiences survey (n = 715 respondents preintervention, 437 respondents postintervention). We compared pilot practices' performance on utilization and patient survey measures to 18 comparison practices using difference-in-difference regression models accounting for practice-level clustering. We assessed practices' performance on 31 quality measures from the American Society of Clinical Oncology Quality Oncology Practice Initiative program. Practices reported on care for 24 to 40 patients (depending on practice size). For aim 2, we conducted site visits to interview providers and patients and observe clinical encounters. External reviewers evaluated practices' implementation of the standards.
There were no statistically significant differences in hospitalizations, ED use or primary care visits between the pilot and comparison group over time, but there was a significant increase in specialty provider visits (includes oncology and other specialties) for the pilot group. Unadjusted rates of specialty visits were 0.92 PMPM for the pilot group and 0.91 for comparison group at baseline; at follow-up the rates were 1.21 and 1.13 respectively. The adjusted difference-in-difference of 0.07 (CI, 0.01-0.13) was significant ( = .03). For the patient experience survey, we calculated 5 composite scores and an overall rating score on a 0 to 100 scale. Survey results showed that pilot practices improved more than comparison practices on shared decision-making (4.03 difference-in-difference composite score, = .013), while the comparison group had greater improvement on 2 composites: Access (−6.36 difference-indifference composite score, < .001) and Exchanging Information (−4.25 difference-indifference composite score, = .013). There were no statistically significant differences on other measures. On average, the pilot practices improved performance on 65% of core quality measures and 73% of palliative care measures from baseline to follow-up; the greatest improvements were in symptom assessment and care planning. Qualitative analyses revealed that practices focused quality improvement on 5 areas: (1) goals of care, (2) engaging patients in quality improvement, (3) financial counseling, (4) symptom management, and (5) care coordination.
Oncology practices that implemented the patient-centered oncology standards in this pilot demonstrated improved quality on several patient-centered processes. However, there was no impact on hospitalizations/ED use and mixed results for patient experiences for pilot practices relative to comparison practices. This was the first intervention to be evaluated in oncology practices, and findings are consistent with early evaluations of primary care medical homes.
Patient attribution to pilot and comparison practices was imperfect as patterns of attribution changed over time and some pilot practices were underrepresented in the utilization and survey samples. Also, we do not have information on comparison practices' care systems or quality. The study was not designed to analyze the effect of the standards on subpopulations of practices or patients.
