Introduction
Although adverse social determinants of health (aSDoH) are associated with poorer health outcomes,1 and are highly prevalent in ED patients,2 many screening efforts have only been in primary care settings. Primary care based screening may exacerbate disparities, as patients with barriers to accessing primary care never receive resources. We therefore sought to describe the demographic characteristics and ED utilization of Medicaid patients who were missed by a primary care screening program.
Methods
We conducted a retrospective cohort study using data from a large Medicaid Accountable Care Organization (ACO) in Massachusetts. Patients were eligible if they had at least 11 months3 of enrollment between February 2019-February 2020, during the initial aSDoH screening period.
We linked two data sources: (1) primary care aSDoH questionnaire and (2) Medicaid ACO claims data. Patients could screen positive for aSDoH by affirming a risk (social risk) or by asking for information (social need). Data on demographics (age, sex, race, ethnicity, primary language) was acquired from the electronic medical record system. Eligible primary care provider (PCP) visits were defined as a visit to an in-system PCP that participated in screening (Methods Appendix). Patients were defined as “screened” if they partially completed at least one screen during the study period and defined as “missed screening” if they had no completed screens, regardless of their visit status or reason for not having completed the screening (e.g. refusal v. not being offered). For members with >1 screen during the study period we used the most positive screening result.
We described ED utilization during the study period for patients by screening status using Stata SE 15.1 (StataCorp, College Station, TX). The study was determined to be exempt by the MGB IRB.
Results
There were 77,524 Medicaid ACO members in the study cohort; 38,370 (49%) adults (18–65 years), 27,848 (36%) children (5–17 years), and 11,304 (15%) infants/toddlers (0–4 years) (Table 1). In the cohort, 43,199 (56%) were female. 29,316 (38%) were non-Hispanic White, 6,556 (8%) were non-Hispanic Black, 23,916 (31%) were Hispanic. 61,412 (79%) reported a primary language of English, and 10,320 (13%) reported Spanish. A total of 59,803 (77%) had a visit to a PCP providing screening.
Table 1.
Demographic characteristics of members stratified by screening status (e.g., missed vs. screened).
| Overall (all members) | Missed Screening | Screened | |||||
|---|---|---|---|---|---|---|---|
| Missed overall | No PCP visit | PCP visit and not screened | Screened overall | Screened negative (no social risk or need present) | Screened positive (social risk or need present) | ||
| Total n (%) | 77,524 | 50,753 (65) | 16,267 (32) | 34,486 (68) | 26,771 (35) | 13,199 (49) | 13,572 (51) |
| Age categories, n (col %) | |||||||
| 0–4 years | 11,304 (15) | 5,518 (11) | 624 (4) | 4,894 (14) | 5,786 (22) | 2,852 (22) | 2,934 (22) |
| 5–17 years | 27,848 (36) | 18,489 (36) | 3,654 (22) | 14,835 (43) | 9,359 (35) | 5,095 (39) | 4,264 (31) |
| 18–65 years | 38,370 (49) | 26,745 (53) | 11,989 (74) | 14,756 (43) | 11,625 (43) | 5,252 (40) | 6,373 (47) |
| Sex, n (col %) | |||||||
| Female | 43,199 (56) | 27,915 (55) | 8,668 (53) | 19,247 (56) | 15,284 (57) | 7,441 (56) | 7,843 (58) |
| Male | 34,324 (44) | 22,837 (45) | 7,598 (47) | 15,239 (44) | 11,487 (43) | 5,758 (44) | 5,729 (42) |
| Missing | 1 | 0 | 1 | 0 | 0 | 0 | 0 |
| Race/ethnicity, n (col %) | |||||||
| Non-Hispanic White | 29,316 (38) | 19,297 (38) | 6,509 (40) | 12,788 (37) | 10,019 (37) | 5,477 (42) | 4,542 (33) |
| Non-Hispanic Black | 6,556 (8) | 4,164 (8) | 1,390 (9) | 2,774 (8) | 2,392 (9) | 980 (7) | 1,412 (10) |
| Hispanic | 23,916 (31) | 14,986 (30) | 4,389 (27) | 10,597 (31) | 8,930 (33) | 3,855 (29) | 5,075 (37) |
| Non-Hispanic other | 2.633 (3) | 1,662 (3) | 470 (3) | 1,192 (3) | 971 (4) | 501 (4) | 470 (3) |
| Non-Hispanic unknown | 3,355 (4) | 2,152 (4) | 648 (4) | 1,504 (4) | 1,203 (4) | 586 (4) | 617 (5) |
| Missing | 11,748 (15) | 8,492 (17) | 2,861 (18) | 5,631 (16) | 3,256 (12) | 1,800 (14) | 1,456 (11) |
| Language preference, n (col %) | |||||||
| English | 61,412 (79) | 40,036 (79) | 13,184 (81) | 26,852 (78) | 21,376 (80) | 10,800 (82) | 10,576 (78) |
| Spanish | 10,320 (13) | 6,525 (13) | 1,497 (9) | 5,028 (15) | 3,795 (14) | 1,596 (12) | 2,199 (16) |
| Other | 3,152 (4) | 2,030 (4) | 484 (3) | 1,546 (5) | 1,122 (4) | 530 (4) | 592 (4) |
| Missing | 2,640 (3) | 2,162 (4) | 1,102 (7) | 1,060 (3) | 478 (2) | 273 (2) | 205 (2) |
Abbreviations: PCP, primary care practice.
In the cohort, 26,771/77,524 (35%) ACO members were screened and 50,753/77,524 (65%) were missed. Of the missed patients, 16,267/50,753 (32%) had no documented visit to a screening PCP.
Of those who missed screening (n=50,753), 10,005/26,745 adults (37%) and 7,160/24,007 (30%) children had an ED visit; 5,461 adult visits and 2,355 pediatric visits were to in-system EDs (Figure). Among adults and children who used an in-system ED, 60% (7,816/13,038) were missed in primary care screening. Rates of ED visits before and after eligible PCP visits by screening status are shown in the Appendix Table.
Figure:
ED Utilization by Screening Status
Discussion
In this cohort, 34% of individuals with missed primary care aSDoH screening presented to an ED during the 1-year study period, with 15% presenting to an ED within the health system. Though drawn from the early implementation process, these data suggest that improved attention to screening implementation in clinics is needed and that an ED-inclusive strategy may provide the opportunity to capture patients who were missed.
Limitations include the inclusion of only consistently-enrolled Medicaid members, and the fact that screening rates may have improved over time. Although the screening rates in this cohort are similar to other studies,5 there is potential misclassification. For example, 1,455 patients were screened without an eligible visit, suggesting that some PCPs were using screening that we did not capture. Additionally, we are unable to identify visits to PCPs who were not screening.
Including the ED in a population-health screening strategy may help to avoid exclusion, and patients are supportive of ED-based screening.6 To do so, efficient and effective ED strategies are needed, potentially including multi-lingual screening tools sent to patients’ own devices and asynchronous navigation provided after an ED visit.
Approximately half of ED visits were to EDs outside of our healthcare system, suggesting that tools to allow data sharing between health systems would help to enhance screening and improve connection with resources. These could include EMR alerts for patients who have missed screening or have positive screening results that have not yet been addressed. ACOs should consider incentivizing ED-inclusive screening pathways to reduce inequities, although rates of ED utilization were similar regardless of screening status, suggesting that screening alone is insufficient to reduce ED utilization.
Supplementary Material
Acknowledgements
This work was supported by a grant from the Society for Academic Emergency Medicine Foundation. Dr. Thorndike was supported by NIH grant K24 HL163073.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Bensken WP, Alberti PM, Koroukian SM. Health-Related Social Needs and Increased Readmission Rates: Findings from the Nationwide Readmissions Database. J Gen Intern Med 2021; 36(5): 1173–80. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Malecha PW, Williams JH, Kunzler NM, Goldfrank LR, Alter HJ, Doran KM. Material needs of emergency department patients: A systematic review. Acad Emerg Med 2018; 25(3): 330–59. [DOI] [PubMed] [Google Scholar]
- 3.Peltz A, Samuels-Kalow ME, Rodean J, et al. Characteristics of Children Enrolled in Medicaid With High-Frequency Emergency Department Use. Pediatrics 2017; 140(3): [Epub]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Samuels-Kalow ME, Faridi MK, Espinola JA, Klig JE, Camargo CA Jr. Comparing Statewide and Single-center Data to Predict High-frequency Emergency Department Utilization Among Patients With Asthma Exacerbation. Acad Emerg Med 2017; 25(6): 657–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Schickedanz A, Sharp A, Hu YR, et al. Impact of social needs navigation on utilization among high utilizers in a large integrated health system: a quasi-experimental study. J Gen Intern Med 2019; 24(11): 2382–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Samuels-Kalow ME, Molina MF, Ciccolo GE, et al. Patient and Community Organization Perspectives on Accessing Social Resources from the Emergency Department: A Qualitative Study. West J Emerg Med 2020; 21(4): 964–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.