Dear Editor,
Over 500,000 people experience homelessness in the United States on any given night.1 Skin disease is prevalent in homeless individuals, and increased morbidity has been attributed to environmental exposures, lack of resource access, and inconsistent healthcare.1,2 In this case-control study, we identify differences in dermatologists’ management of common conditions in homeless relative to non-homeless patients that may contribute to healthcare disparities in the homeless.
In Salt Lake City, Utah, free primary and specialty care is provided by Fourth Street Clinic (FSC) to homeless persons. Since 2000, dermatologists from the University of Utah (UU) Department of Dermatology voluntarily staff FSC’s twice-monthly referral-based dermatology clinic. Homeless FSC patients have similar access to medications, procedures, and follow-up as UU clinics, albeit free-of-charge. Management practices of seven UU dermatologists who volunteer regularly at FSC (attending ≥seven clinics between 2011–2017) were analysed.
Adult (>17 years) homeless (FSC clinics) and non-homeless patients (UU clinics) diagnosed with actinic keratosis, dermatitis, non-genital verrucae, or acne between May 2011 and July 2017 were included. Demographic and treatment information were obtained via electronic health records. Five non-homeless patients were matched to each homeless patient based on diagnosis, gender, age ±3 years, treating dermatologist, and clinic visit date ±3 years using Greedy matching (SAS macro %gmatch).3 Matched-control patients (n=98) with duplicate entries, no clinic note, non-matching diagnosis, or no evidence of skin disease were excluded.
Dermatitis diagnoses included atopic, seborrheic, contact, and not-otherwise-specified subtypes. Procedures for dermatitis included biopsy, potassium hydroxide preparation, intralesional steroid injection, and culture. Acne severity was determined using the Global Acne Grading System.4 We compared homeless to non-homeless matched-control patients using t-tests, Kruskal-Wallis tests, or chi-squared tests, as appropriate. Analyses were performed in SAS version 9.4 (SAS, Cary, NC).
Fifty-seven homeless patients with AKs were matched to 283 non-homeless patients and had comparable numbers of AKs (p=0.25). Overall, medications were prescribed in 7.4%, and cryotherapy was performed in 89.4% of patients. Comparing homeless and non-homeless patients, no significant differences existed between the number of prescriptions or treatment with cryotherapy (p=0.51 and p=0.06, respectively, Table 1). However, stratifying by number of AKs, in those with few (<5) AKs, cryotherapy was more common for non-homeless compared to homeless patients (93.8% vs. 81.3%, p=0.02). In addition, follow-up was more commonly recommended for non-homeless patients (86.9% vs. 36.8%, p<0.001).
Table 1.
Comparison of treatment for common skin conditions, overall and by homeless status.
| Overall | Non-Homeless | Homeless | P-Value | ||
|---|---|---|---|---|---|
| Actinic Keratosis, N | 340 | 283 | 57 | ||
| Prescription Given, N (%) | No | 315 (92.6) | 261 (92.2) | 54 (94.7) | 0.51 |
| Yes | 25 (7.4) | 22 (7.8) | 3 (5.3) | ||
| Cryotherapy, N (%) | No | 36 (10.6) | 26 (9.2) | 10 (17.5) | 0.06 |
| Yes | 304 (89.4) | 257 (90.8) | 47 (82.5) | ||
| Recommended Follow Up, N (%) | No | 73 (21.5) | 37 (13.1) | 36 (63.2) | <0.001 |
| Yes | 267 (78.5) | 246 (86.9) | 21 (36.8) | ||
| Dermatitis, N | 303 | 241 | 62 | ||
| Strongest Potency Steroid | |||||
| Prescribed, N (%) | High | 112 (37.0) | 101 (41.9) | 11 (17.7) | <0.001 |
| Medium | 133 (43.9) | 90 (37.3) | 43 (69.4) | ||
| Low | 26 (8.6) | 20 (8.3) | 6 (9.7) | ||
| None | 32 (10.6) | 30 (12.4) | 2 (3.2) | ||
| Procedure Performed, N (%) | No | 249 (82.2) | 188 (78.0) | 61 (98.4) | <0.001 |
| Yes | 54 (17.8) | 53 (22.0) | 1 (1.6) | ||
| Recommended Follow-Up, N (%) | No | 193 (63.7) | 147 (61.0) | 46 (74.2) | 0.054 |
| Yes | 110 (36.3) | 94 (39.0) | 16 (25.8) | ||
| Non-genital verrucae, N | 227 | 199 | 28 | ||
| Prescription Given, N (%) | No | 167 (73.6) | 141 (70.9) | 26 (92.9) | 0.01 |
| Yes | 60 (26.4) | 58 (29.1) | 2 (7.1) | ||
| Cryotherapy, N (%) | No | 43 (18.9) | 40 (20.1) | 3 (10.7) | 0.24 |
| Yes | 184 (81.1) | 159 (79.9) | 25 (89.3) | ||
| Recommended Follow-Up, N (%) | No | 139 (61.2) | 117 (58.8) | 22 (78.6) | 0.04 |
| Yes | 88 (38.8) | 82 (41.2) | 6 (21.4) | ||
| Acne, N | 153 | 126 | 27 | ||
| Total # of Prescriptions Per Patient, N (%) | 0.005 | ||||
| 0 | 2 (1.3) | 2 (1.6) | 0 (0.0) | ||
| 1–2 | 107 (69.9) | 82 (65.1) | 25 (92.5) | ||
| 109 (71.3) | 84 (66.7) | 25 (92.5) | |||
| 3+ | 44 (28.7) | 42 (33.3) | 2 (7.4) | ||
| Recommended Follow Up, N (%). | No | 48 (31.4) | 33 (26.2) | 15 (55.6) | 0.003 |
| Yes | 105 (68.6) | 93 (73.8) | 12 (44.4) |
Sixty-two homeless patients with dermatitis were matched to 241 non-homeless patients with comparable numbers of body sites affected (Average: 2; p=0.73); 96.8% of patients were prescribed medications. High-potency topical steroids were less commonly prescribed in homeless patients (17.7% vs. 41.9%), whereas mid-potency topical steroids were more frequently prescribed (69.4% vs. 37.3%, p<0.001). Procedures were performed less often in homeless patients (1.6% vs. 22.0%, p<0.001), notably, biopsies (0% vs. 10.8%, p=0.007). Although follow-up was recommended less frequently for homeless patients, this did not reach statistical significance (25.8% vs. 39.0%, p=0.054).
Forty-five homeless patients with verrucae matched to 207 non-homeless patients did not differ in verruca counts (Average: 1; p=0.33). Cryotherapy was performed in 89.3% of homeless and 79.9% of non-homeless patients. Homeless patients were less commonly prescribed medications (7.1% vs. 29.1%, p=0.01) and were less commonly recommended follow-up (21.4% vs. 41.2%, p=0.04).
Twenty-seven homeless patients with acne were matched to 126 non-homeless patients. No differences existed in distribution of acne severity (p=0.64). Homeless patients were more commonly prescribed ≥1 topical medications compared to non-homeless patients (92.6% vs. 72.2%, p=0.001). Non-homeless patients were prescribed more acne medications at the same visit (more complicated regimen) compared to homeless patients (p=0.03). Follow-up was less frequently recommended for homeless patients (44.4% vs. 73.4%, p=0.003).
This study highlights patterns in less diagnostic inquiry, less aggressive intervention, and fewer recommendations for follow-up for common skin conditions in homeless populations versus non-homless populations, despite the same dermatologists treating both groups. As we reported previously, these trends may contribute to disparities associated with homelessness.5 Potential explanations for these differing dermatologist behaviors may include lack of knowledge of available resources at the homeless clinic, implicit biases, concern for patient adherence or resource stewardship, or tendency to over-medicalize non-homeless patients.6,7 Adopting available guidelines for treating common skin conditions or improving physician training about homelessness and its effects may reduce treatment disparities.
Although limited by sample size, single geographic location, and potential confounding comparisons between the homeless community clinic and “non-homeless” academic medical center clinics, our findings demonstrate that healthcare disparities exist in the dermatologic care given to homeless compared to non-homeless patients. This is particularly apparent in dermatologists’ diagnostic and treatment approaches towards common skin conditions. Underlying causes for these disparities need to be explored and mitigated.
Acknowledgements:
We would like to thank Tiffiny Gregory, Will Lowder, and Mingyuan Zhang for data extraction support.
Funding sources: Dr. Secrest receives support from a Dermatology Foundation Public Health Career Development Award. Ms. Truong is supported by a National Institutes of Health predoctoral fellowship, F30CA235964.
Footnotes
Conflict of Interest: None to declare
References
- 1.Henry MMA, Morrill T, Robinson A, et al. The 2018 annual homeless assessment report (AHAR) to Congress. Part 1: Point-in-time estimates of homelessness. In: The U.S Department of Housing and Urban Development, ed 2018. [Google Scholar]
- 2.Stratigos AJ, Katsambas AD. Medical and cutaneous disorders associated with homelessness. Skinmed. 2003;2(3):168–172; quiz 173–164. [DOI] [PubMed] [Google Scholar]
- 3.Kosanke J and Bergstralh E. GMATCH: A computerized matching of cases to controls using the greeding matching algorithmwith a fixed number of controls per case. 2004. http://bioinformaticstools.mayo.edu/research/gmatch/
- 4.Doshi A, Zaheer A, Stiller MJ. A comparison of current acne grading systems and proposal of a novel system. Int J Dermatol. 1997; 36(6): 416–8. [DOI] [PubMed] [Google Scholar]
- 5.Truong A, Laggis CW, Annis TD, et al. Factors associated with follow-up adherence in patients seen at a referral-based clinic for the homeless. J Am Acad Dermatol. 2019. October 8; ePub ahead of print. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Baggett TP, Liauw SS, Hwang SW. Cardiovascular disease and homelessness. JACC. 2018;71(22): 2585–97. [DOI] [PubMed] [Google Scholar]
- 7.Stajduhar KI, Mollison A, Giesbrecht M, et al. “Just too busy living in the moment and surviving”: barriers to accessing health care for structurally vulnerable populations at end-of-life. BMC Palliative Care. 2019;18(11): 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]