Study 1
Wang and colleagues (2012) conducted a randomized controlled trial to study the effect of primary care–based, complex care management as an alternative to expedited emergency primary care for chronically ill individuals who were released from prison within the past 2 weeks. The study was in San Francisco, Calif.
The treatment group was offered an alternative to accessible primary care, called Transitions Clinic (TC). Participants who initially enrolled received an expedited primary care appointment, but future appointments were made at the discretion of the TC provider. However, patients could call the provider for urgent medical issues. The comparison group received an expedited primary care appointment within 4 weeks with a primary care–clinic provider who did not have formal training for previous prisoner populations. The comparison group also had access to the same services, but was not assisted by a community health worker.
To be eligible for the study, individuals had to 1) be English speaking; 2) be 50 years or older; 3) have at least one chronic illness, including mental health conditions or addiction; and 4) not have a primary care provider in San Francisco. The treatment group participants (n=98) had an average age of 42.9 years, Of this group, 23.5 percent were married, and 91.8 percent were male. The race/ethnicity of the treatment group was 63.4 percent Black, 20.8 percent white, 10.4 percent Hispanic, and 5.4 percent Asian. One third of the group was insured, 13 were covered by Medicaid, 4 were covered by Medicare, and 4 were covered by Healthy San Francisco. The comparison group participants (n=102) had an average age of 43.6, 13.7 percent were married, and 96.5 percent were male. The race/ethnicity of the comparison group was 65 percent Black, 17 percent white, 14 percent Hispanic, and 3 percent Asian. Four percent were employed, while 29.4 percent were insured, 11 were covered by Medicaid, 4 were covered by Medicare, and 3 were covered by Healthy San Francisco. There was a statistical adjustment for time incarcerated and deaths; however, the rest of the baseline data was determined to be statistically equivalent.
The baseline data was collected on a participant questionnaire. Participants self-rated health questions and diagnoses, past health care while incarcerated and before incarceration, case worker interaction, and past hospitalization history. Information on age, gender, race, and ethnicity was collected from the electronic data repository. The two primary health care outcomes that were assessed were 1) two or more visits to the study-assigned, primary care clinic; and 2) visits to the medical or psychiatric emergency room that did not result in hospitalization. Two secondary health care outcomes, rate of emergency room use and of any hospitalization, were also assessed. Finally, two incarceration outcomes were also assessed: 1) first-time incarceration, and 2) time to first incarceration, which included short stays not resulting in conviction and longer, stays resulting in transfer to the state prison system.
An intent-to-treat analysis was conducted, using a chi-square test to compare primary care, emergency room use, hospitalizations, and incarceration. The Wilcoxon rank sum test was used to compare emergency room visits, and applied Poisson regression was used to compare the rates of emergency room utilization and hospitalization. The Poisson regression adjusted for the time incarcerated and death. The follow-up period was 12 months.