Primary Care Clinical Excellence Incentive Study (PCCE)

As part of UCLA Health's commitment to developing a premier integrated health system built on a foundation of physician-led, team-based primary care, the Department of Medicine (DOM) recently implemented a new performance based incentive plan called the Primary Care Clinical Excellence (PCCE) Incentive Plan. This incentive plan was developed to motivate providers to improve health maintenance screening rates. The UCLA Health DOM Quality team is leading the implementation and evaluation of this new incentive plan across our primary care network. In addition, the DOM Quality team has partnered with the UCLA Anderson School of Management to study the most efficacious ways to frame and communicate performance based incentives. Understanding the factors that motivate physicians to deliver the highest quality primary care will provide pivotal insights into the successful implementation of performance based programs nationwide. The investigators believe that physicians who receive communication built on behavioral principles will demonstrate more motivation towards and success at meeting national primary care screening guidelines.

Currently, UCLA Health has 42 primary care clinics staffed with approximately 205 PCP's serving more than 300,000 patients. In order to improve health maintenance screening rates associated with the performance based program, the investigators will use a communication strategy that leverages behavioral principles to motivate providers to improve health maintenance screening rates. The investigators will implement a three-arm experimental communication campaign that includes monthly messages. The communication strategy will utilize behavioral change theory (particularly social comparison and recognition) to improve physician performance in health maintenance quality metrics. This study will primarily focus on one quality domain of the PCCE incentive program-Clinical Quality. Two further quality domains of the PCCE incentive program will serve as exploratory measures: Patient Experience and Professional Participation. For the primary outcome measure and high-priority secondary outcome measures, the investigators will first evaluate whether the combination of arms 2 and 3 differs from arm 1 in terms of the measures listed in the Outcome Measures section. If this comparison is statistically significant, the investigators will next compare arms 1 vs 2. and then, if also significant, the investigators will compare arms 1 vs 3 and 2 vs 3. For the 1 vs 3 and 2 vs 3 comparisons, we will use a Holm-Bonferroni p-value correction. These comparisons of treatment arms will be performed at the patient level using a mixed effects regression model, including random physician and clinic effects to account for clustering of patients. Control variables include (1) patient baseline order rates (or completion rates depending on the outcome measure) in July-October 2019, (2) patient characteristics (age, gender, comorbidity, insurance plan, zip code), and (3) physician characteristics (gender, race, years practiced, years at UCLA Health). Unless otherwise specified, p-values less than 0.05 will be considered statistically significant.

Behavioral Economics, Primary Health Care, Health Maintenance, Diabetes, Colorectal Cancer Screening, Cervical Cancer Screening, Chlamydia Screening, HPV Screening

Parallel cluster randomized trial. Groups will be clustered at the clinic level taking into consideration total Clinical FTE, Primary Care network group, and baseline HM completion rates.

Monthly email communication: Monthly standard communication via email informing physicians of their health maintenance completion rate over the prior three-month period. The email will also include a link to the performance dashboard, a link to a FAQ page that outlines their incentive plan, a link to helpful resources such as care guidelines, key tips from top performers, the quality measure on which they are performing the best, and the two quality measures that they could improve on the most.

Monthly email communication as in Arm 1 Social comparison: In addition to the information given in Arm 1 email messaging, physicians in this arm will receive a list of the names of the top 25 performers from the prior month, and a high performer benchmark. Each physician in this group will receive a personalized message and subject line based on where in the performance distribution they fall (categories: top 25 performers, high performers, nearly high performers, and low performers). Message content will incorporate language utilizing principles of social comparison theory.

Monthly email communication as in Arm 1 and Arm 2. Social comparison as in Arm 2. Quality improvement leadership training: Physician leads and clinic managers within clinics randomized to Group 3 will receive an in-person quality improvement and primary care clinic performance training, using the principles of quality improvement, self-determination theory and social comparison. Clinic leaders will be trained on how to guide these conversations, formulate their own performance/quality improvement goals, design effective strategies to reach these goals, and track their clinic's progress. Check-in emails and calls will be provided on a monthly basis to follow up on clinic-based quality improvement efforts and share key take-aways from the new Primary Care Clinical Excellence Recognition Program. This program aims to support a positive and collaborative culture of clinical excellence by recognizing and sharing best practices from high performing physicians and clinical teams.

The social comparison is added onto the standard communication email to provide some peer comparison between the physicians.

Clinic and Physician leads will be provided with a clinic leadership training seminars in leadership, quality improvement, and support from the Quality Improvement team at UCLA Health.

For each patient who visits a given PCP in our sample from November 5, 2019 to March 3, 2020, we will consider all Focus Quality Measures open at his/her first visit to the PCP, and calculate the percent of these measures ordered within 7 days following his/her first visit. Focus Quality Measures include: Diabetes HGB A1C, Diabetes Eye Exam, Diabetes Foot Exam, Diabetes Nephropathy Monitoring, Colon Cancer Screening, Breast Cancer screening, Cervical Cancer Screening, chlamydia screening, and HPV depending on primary care specialty. These Focus Measures are based on the USPSTF primary care recommendations. [Note: The original intent was to run the study over an eight month period - from November 1, 2019 through June 30, 2020 - but was cut short due to the COVID-19 pandemic. Our revised time frame includes (1) four full cycles (months) of the intervention (11/05/2019-03/03/2020) and four months prior to the experiment as a baseline period]

For each patient who visits a given PCP in our sample from November 5, 2019 to March 3, 2020, we will consider all Focus Quality Measures that are open between November 5, 2019 and March 3, 2020. We will calculate the percentage of these measures completed between November 5, 2019 and March 10, 2020 as well as the percentage completed between November 5, 2019 and July 3, 2020.

For each patient who visits a given PCP in our sample from November 5, 2019 to March 3, 2020, for each Focus Quality Measure open at his/her first visit to the PCP, we will assess whether the open measure is ordered within 7 days following his/her first visit.

For each patient who visits a given PCP in our sample from November 5, 2019 to March 3, 2020, for each Focus Quality Measure open between November 5, 2019 to March 3, 2020, we will assess whether the measure is completed (1) between November 5, 2019 and March 10, 2020 as well as (2) between November 5, 2019 and July 3, 2020.

As measured by aggregate order rates for Complementary Health Maintenance Measures within seven days after each patient's first visit. Complementary Measures (Abdominal Aortic Aneurysm Screening, Osteoporosis Screening, Hepatitis C Screening, HIV Screening, Pneumococcal Vaccination, Tetanus Vaccination, Tdap Vaccination During Pregnancy, Shingles Vaccination, Meningitis Vaccination, Measles Mumps Rubella Vaccination, Polio Vaccination, Hepatitis A Vaccination, Hepatitis B vaccination, Pneumococcal Vaccination for Patients with Diabetes, Statin for Primary ASCVD Prevention, Aspirin for Secondary ASCVD Prevention, and Annual Wellness Visit) are based on USPSTF primary care recommendations regarding disease management, infectious disease screening, and standard vaccinations.

As measured by patient level aggregate performance Complementary Health Maintenance Measures (1) between November 5, 2019 and March 10, 2020 as well as (2) between November 5, 2019 and July 3, 2020.

Physician performance in professional participation will be based on provider/physician surveys throughout the study period.

For each patient who visits a given PCP in our sample from November 5, 2019 to March 3, 2020, we will consider all Focus Quality Measures that are open between November 5, 2019 and March 3, 2020, and calculate the percentage of these measures ordered between November 5, 2019 and March 10, 2020.

HM order rate is determined by the aggregate order rate of measures within the primary care Health Maintenance module of EPIC EHR ("Focus Quality Measures") and other Quality Measures that are not in the Health Maintenance module ("Complementary Measures"). Both the Focus Quality Measures and Complementary Measures are based on USPSTF primary care recommendations regarding disease management, infectious disease screening, cancer screening, and standard vaccinations.

HM completion rate is determined by the aggregate completion rate of variables within the primary care Health Maintenance module of EPIC EHR ("Focus Quality Measures") and other Quality Measures that are not in the Health Maintenance module ("Complementary Measures").

For each patient who visits a given PCP in our sample from November 5, 2019 to March 3, 2020, for each Focus Quality Measure open between November 5, 2019 to March 3, 2020, we will assess whether the measure is ordered between November 5, 2019 and March 10, 2020.

For each primary care physician, we assign patients to their panel using an attribution model. This model assigns patients to a physician if they have been seen by that physician within the last 3 years. Additional patient attribution logic includes: 1) the physician with a preventive/wellness visit in the prior 1 year is attributed first, 2) if there is no preventive/wellness visit in the prior 1 year, the physician with the largest number of visits is attributed, and 3) if there is a tie in either the preventive/wellness visit or number of visit scenario, the physician with the most recent visit is attributed. Once patients are assigned using the above attribution model, patients who have at least one visit with their attributed physician from November 5, 2019 to March 3, 2020 will be included in our evaluation.

For each physician, we will identify the number of new patients they have from November 5, 2019 to March 3, 2020.

For each patient who visits a given PCP in our sample from November 5, 2019 to March 3, 2020, we will measure the total number of times he/she visits the PCP from November 5, 2019 to March 3, 2020.

For each patient who visits a given PCP in our sample from November 5, 2019 to March 3, 2020, we will measure the number of open measures he/she has at the first visit as well as the total number of open measures he/she has from November 5, 2019 to March 3, 2020.

Inclusion Criteria Primary care physicians within the UCLA Health Department of Medicine Primary Care Network and with the following specialties: Internal Medicine, Geriatrics, Internal Medicine/Pediatrics, or Family Medicine. Clinical full-time employee level (FTE) of ≥ 40% at the beginning of the intervention period. Eligible to receive performance-based financial incentive as determined by the Department of Medicine. Panel size >50 patients before the first intervention email was sent Patients, with at least one Health Maintenance topic open before or during at least one visit that takes place with a physician enrolled in the study between November 5, 2019 and March 3, 2020. Exclusion Criteria: Not a primary care physician within the UCLA Health Department of Medicine Primary Care Network. Not a primary care physician with the following specialties: Internal Medicine, Geriatrics, Internal Medicine/Pediatrics, or Family Medicine. Clinical FTE < 40%. Not eligible to receive performance-based financial incentive as determined by the Department of Medicine. Panel size greater than 50 patients before the first intervention email was sent. Patients that are not seen by a provider enrolled in the study period. Patients seen by physician in the study with no Health Maintenance topics open before or during any visit that takes place between November 5, 2019 and March 3, 2020.