Role of Psychiatric Hospitals in the Equitable Distribution of COVID-19 Vaccines

We analyzed publicly available data from the Inpatient Psychiatric Facility Quality Reporting (IPFQR) program (4) (federal fiscal year 2018), which collects data from nearly all U.S. IPFs. We analyzed three program measures relevant to the research questions.

The influenza immunization (IMM-2) measure represents the proportion of patients screened for seasonal influenza immunization status and vaccinated before discharge, if indicated, among all inpatients discharged from October 1 to March 31 (influenza season) of the measurement year. This measure captures two activities: screening and vaccination, when indicated. As a result, the IPF receives credit for the immunization measure even if the patient had documented contraindications, declined, or had already received the vaccine during the current year’s influenza season (5).

The timely transmission of transition record (TTR) measure (6) represents the proportion of patients discharged from an IPF for whom a transition record was sent to another facility, primary care provider, or other health care professional designated for follow-up care within 24 hours of patient discharge. The transition record must contain a core set of elements related to the patient’s diagnosis, treatment, and care plan. Transmission can take place by fax, secure e-mail, or mutually accessible electronic health records.

The follow-up after hospitalization for mental illness (FUH-30) measure (7) represents the proportion of Medicare beneficiaries discharged from IPFs who received follow-up care from a mental health provider within 30 days.

Each facility receives a single score for each measure, representing the proportion of discharged patients who met the requirements. Higher scores indicate better performance. For the current study, in addition to examining performance on the IMM-2, we analyzed performance on record transmission and follow-up, because these measures provide insight into the extent to which community providers would have the information required to ensure that patients discharged from an IPF receive their second COVID-19 vaccine dose and the extent to which these patients remain engaged with the health care system immediately after discharge. IPFs report their IMM-2 and TTR through their health records or other administrative data, whereas CMS calculates the follow-up scores on behalf of IPFs using Medicare data. Details about the measure specifications and populations included are available (4). The reporting program includes IPFs in every state and Washington, D.C. We excluded eight IPFs in Puerto Rico for these analyses.

We used the IPFs’ addresses to merge the IPFQR data with county-level COVID-19 Community Vulnerability Index data from the Surgo Foundation (8). The vulnerability index draws on multisource data to rank the COVID-19 vulnerability of geographic areas (state, county, or census tract) relative to other geographic areas, across quintiles of very low, low, moderate, high, and very high vulnerability. Each geographic area receives a score from 0 to 1; higher scores indicate higher vulnerability. The vulnerability index captures 34 factors that reflect community socioeconomic status, household composition, disability, racial-ethnic minority status, language, house type, availability of transportation, epidemiologic factors (including high-risk COVID-19 populations), and health system strength and capacity. We conducted descriptive analyses to examine the distribution of IPFQR measure scores and the relationship between those scores and county-level COVID vulnerability scores.

Of the 1,602 IPFs included in our analytic file, nearly all had scores for the IPFQR measures and were included in this study. On average, 84% of patients were screened for influenza vaccination status and were vaccinated if indicated (range 0%–100% across IPFs; median=94%). According to the TTR scores for calendar year 2018, the IPFs reported 961,730 discharges. On average, 57% of patients had their records transmitted to another provider within 24 hours of discharge (range 0%–100% across IPFs; median=67%), and half of Medicare patients had a follow-up visit with a mental health provider within 30 days of discharge (range 5%–96% across IPFs; median performance=50%). Across all measures, the lowest-performing facilities (defined as scoring below the first quartile) were concentrated in the South.

Twenty-three percent of IPFs were in counties with high or very high COVID-19 vulnerability index scores (Table 1). Average IMM-2 scores were similar across counties with different levels of COVID-19 vulnerability, but counties with higher COVID-19 vulnerability had better average TTR scores but worse average FUH-30 scores. For example, in counties with very high COVID-19 vulnerability, 32% of the facilities had a low score on follow-up care, compared with only 10% of facilities in counties with very low COVID-19 vulnerability. Consistent with these findings, COVID-19 vulnerability scores were correlated with TTR and FUH-30 measure performance (r=0.1 and –0.27, respectively, both p<0.001) but not with performance on the IMM-2 (r=−0.003).

Overall, IPFs have successfully screened patients for influenza immunization status and have provided vaccinations as indicated, suggesting that these facilities could be well positioned to offer COVID-19 vaccines to patients. However, roughly one-quarter of IPFs failed to screen patients for influenza immunization status and provide a vaccination when indicated to at least 80% of their patients.

IPFs would need to develop new protocols for administering COVID-19 vaccines. Unlike influenza vaccinations, some COVID-19 vaccinations require a second dose, which would require facilities to communicate immunization information for discharged patients to the next provider and to ensure that patients receive follow-up care after discharge.

Receipt of follow-up care by discharged patients appeared particularly problematic in communities with high COVID-19 vulnerability, whereas records transmission appeared more problematic in communities with lower COVID-19 vulnerability. We cannot explain the sources of these relationships from the data available, but the findings may help to inform quality improvement efforts. To improve receipt rates for the second vaccine dose, facilities in counties with low COVID-19 vulnerability may need to focus on improving information sharing with the next provider. Facilities in counties with high COVID-19 vulnerability may need to focus on addressing underlying issues that present barriers to receiving follow-up care (e.g., transportation).

To successfully vaccinate the populations they serve, IPFs will likely need to develop stronger relationships with pharmacies and other medical providers. For example, facilities could refer patients to specific pharmacies or community mental health centers that would take responsibility for administering the second vaccine, working in partnership with the IPF to maintain contact with the patient after discharge. Such collaboration would require developing protocols to guide information sharing between providers and tracking of patient care in the community, possibly facilitated by accessible immunization registries. As federal programs and states prioritize COVID-19 vaccinations for patients in long-term and congregate care settings (9), IPFs may also be in the position of administering the second vaccine dose for patients admitted from these settings. To identify best practices to replicate, there would be value in closely monitoring the processes that IPFs and their community partners put into place to support this coordination.

These analyses had several limitations related to the measures. Because of the specifications for the IMM-2, we could not separately report the proportion of patients who received an influenza vaccination and those who declined or had received it before hospitalization. As with an influenza immunization, patients would be able to decline the COVID-19 vaccine or may have received it before their hospital stay. Thus, the IMM-2 is likely the best publicly available indicator for what might happen if these facilities were equipped to administer COVID-19 vaccines. Additionally, the follow-up care measure (FUH-30) captures only follow-up with mental health providers, not primary care providers or pharmacists, both of whom may be more likely than outpatient mental health providers to administer the COVID-19 vaccine. Mental health providers, however, could play a role in reminding patients to obtain their second vaccine dose. For many patients exiting psychiatric facilities, these providers may be the main, or only, point of engagement with the health care system. Community mental health centers have successfully delivered immunizations (2), and some mental health providers offer integrated primary care or have colocated nursing staff qualified to administer vaccines. Finally, IPFs know they are being assessed for these metrics, which may influence their performance. IPFs’ apparent ability to offer and administer influenza immunizations may not translate into success administering COVID-19 vaccines without a similar level of accountability and investment in the clinical and data system infrastructure.

This study was further limited because the county vulnerability index corresponding to the IPF addresses may not completely represent the COVID vulnerability of the communities in which patients live. Facilities that serve large or diverse geographic areas could be able to distribute vaccines to people who would otherwise not have access (e.g., a facility located in a moderately sized city or town may also serve patients who live in rural areas).

Consistent with the recommendations of the National Academy of Medicine’s framework for the equitable distribution of COVID-19 vaccines (10), IPFs could leverage their experience with influenza immunizations to offer patients COVID-19 vaccinations. About one-quarter of these facilities are in communities with high COVID-19 vulnerability. Successfully distributing COVID-19 vaccines in these facilities would require not only focusing efforts on establishing new vaccine distribution processes within the facilities but also on improving transmission of discharge information to other providers and ensuring that patients receive timely follow-up care. As states develop their COVID-19 vaccination plans, they should engage IPFs and the populations they serve to fully assess the factors that could facilitate or impede the successful deployment of vaccines in these facilities.