BY PAUL H. WAYTZ, MD; ANDREW FORSBERG, BA; ABDI MAHAMED, MBAThursday, October 11, 2018
Rheumatoid arthritis is a chronic inflammatory disease that can affect people of all ethnicities. There are no published studies that attempt to elaborate disease-specific characteristics of Somali immigrants. In this paper, we attempt to better categorize various aspects of rheumatoid arthritis for this population, including modes of presentation, laboratory findings, responses to treatment, medication tolerance, and outcomes. A database of 40 Somali patients with rheumatoid arthritis was compiled via chart review, with a goal of examining and analyzing these and other variables. Several cultural differences and other potential barriers emerged, including the possibility of implicit bias. Using the insights from our database, we outline several areas of concern that potentially could be useful to other caregivers in evaluating illnesses in the Somali population, as well as in other ethnic groups.
Rheumatoid arthritis (RA) remains a serious, chronic inflammatory disease with the potential for significant joint damage, as well as systemic morbidity. RA can affect any ethnic group; its prevalence approaches 1% of the population worldwide (1). It is not known if there are variations in clinical presentations and outcomes among various ethnic minorities and in locations where limitations accessing components of medical care may exist. Diagnosis of RA relies primarily on the history of symptoms, as well as a detailed physical examination of joints and attention to other organ systems.
Minnesota is home to the largest Somali population in the world outside of Somalia or refugee camps, with a population now approaching 40,000 residents by conservative estimates (2), the majority of whom live in the Twin Cities. The percentage of Somali who are Muslim is nearly 100%. Most immigrants left Somalia around the time of the country’s civil war in 1991 and over the following 10 years. The current Minnesota population now also reflects a substantial number of second-generation Somali. At first glance, our cohort of RA patients would appear to be a small sample. However, accepting the 1% prevalence, our group of 40 represents 10% of patients in the state, a not insubstantial number for a private medical group. Although sweeping generalizations cannot be made, we feel our observations are worthy of consideration and discussion.
Arthritis and Rheumatology Consultants is a 13-person private practice group located in Edina and Maple Grove, Minnesota. Records of Somali patients were reviewed for demographic and medical data in hopes of discerning a unique cohort that would identify physician and patient variables that might affect both decision-making and outcomes. During this review, it quickly became apparent that socio-cultural factors played an important role in the evaluation and management process, a role that might well be associated with implicit bias. The purpose of this initial study looking into RA in Somali is to examine socio-cultural issues and how various barriers might affect the delivery of care. By extension, we raise the question of how similar barriers might affect medical care of other illnesses in Somali or other ethnic minorities.
Patients were identified based on the needs for a paid interpreter from Jan. 1, 2016 through Dec. 31, 2016. Records of Somali-speaking patients were reviewed for a diagnosis of any rheumatologic complaint based on ICD-9 and/or ICD-10 criteria. A computational spreadsheet was created for data compilation, visual use, and reference. Although patients were identified during the year 2016, charts were mined for dates of first appointments, follow-up appointments, cancelled appointments, and failed appointments (“no-show”). Patients with a diagnosis of RA, either seropositive or seronegative, are those included in this study. Patients with RA met 2010 American College of Rheumatology/European League Against Rheumatism Classification Criteria for RA (3). A number of patients originally diagnosed with inflammatory arthritis at the time of their first appointment were subsequently changed to a diagnosis of RA by their individual clinic provider based on clinical follow-up and supported by laboratory data.
Various social parameters were identified by a standardized printed office questionnaire and clarified by medical assistant personnel via interpreters. In some circumstances, family members were present during appointments along with paid interpreters.
A number of diagnostic tests and therapeutic interventions were obtained during data mining. Results of ESR, CRP, and the autoantibodies rheumatoid factor (RF) and anti-citrullinated peptide (anti-CCP) are reported here to address comparability in our population with previous expectations for RA.
Statistical analysis was performed using the two-tailed T test.
As previously established, all patients were accompanied by a paid interpreter. Sixty-one patients were Somali and, of those, 40 (66%) were identified as having rheumatoid arthritis and form the basis of this study. Other rheumatologic diagnoses included arthralgia (4), Sjogren’s syndrome (3), connective tissue disease (3), polymyalgia rheumatic (3), musculoskeletal pain (3), systemic lupus erythematosus (2), psoriatic arthritis (1), ankylosing spondylitis (1), and giant cell arteritis (1).
Table 1 demonstrates various demographic features of our study group. Our population of Somali with RA is 98% female, distinctly different from the standard reported percentages of nearly 3:1, female to male (4).
Patients’ initial visits involved the typical history, physical examination, and laboratory work-up to best establish a specific diagnosis, exclude other diseases, and determine potential ancillary complicating issues. Our average age of diagnosis was 54.6, perhaps slightly older compared with previously established reports (4). Of note, Table 2 shows that the average age of seropositive patients was 12 years younger than those of seronegative patients, 48.9 versus 60.8 (paired T test, P < .05). A majority of seropositive but no seronegative patients were under the age of 50 when diagnosed. All but two seronegative patients fell between the ages of 50 and 69 (Table 3).
No patients, including the male subject, undressed during the initial appointment. All patients were asked about smoking habits, likely a reflection of the Centers of Medicare and Medicaid Services Meaningful Use mandate of 2011 and implemented in our office in 2012. Only seven patients were asked about alcohol intake and recreational drug use. None had sexual activity histories obtained. Six patients (five women, one man) worked outside of their homes.
Table 3 demonstrates results of markers of inflammation, RF, and anti-CCP antibodies in our Somali population. ESR and CRP results are consistent with previously reported percentages in RA. Both RF and anti-CCP results are somewhat lower than might be expected (5).
Table 4 reports follow-up information in our cohort. This data was pursued based on a previous report from Hennepin County Medical Center (HCMC) that suggested a high rate of failed appointments for follow-up visits. In contrast, our results indicate follow-up to be acceptable and appropriate.
Early diagnosis and treatment of RA are keys to better outcomes, as there appears to be an individual window of opportunity for optimal long-term positive results. The most complete understanding of successful rheumatoid arthritis treatment also involves both sides of the medical partnership and is essential to optimizing decisions and eventual achievement of control. Any impediment to the patient-physician relationship can create confusion, delays, and misunderstandings of otherwise appropriate choices.
Implicit bias refers to unintentional notions and belief patterns that affect our perceptions, behaviors, and decision-making. Implicit as opposed to explicit bias is an unconscious phenomenon that might be uncomfortable to confront and may contradict and conflict with conscious attitudes and values. Furthermore, implicit bias may be just as unacceptable and harm just as much if not more than explicit bias. Recognizing and acknowledging implicit bias when evaluating a Somali patient must be an essential component, as it underlies the essence of caregiving at every step of the process.
Our cohort fit with typical expectations regarding abnormal markers of inflammation and sero- positivity. However, the average age of diagnosis of our patients was slightly older than expected for reasons that may merit further explanation. Of greater interest was that seropositive patients were significantly younger than seronegative patients, an issue that raises several questions—especially if this is unique to Somalis. Though our cohort may be too small to generalize, we regard this as an interesting feature and wonder if this pertains to other ethnic groups with RA.
From the initial moments of an encounter, the presence of an interpreter, even if the interpreter is a family member, creates a barrier. Interpreters are well-educated and trained but may still not know or understand specific rheumatologic terms and questions. Caregivers may have concerns that their explanation of a disease or planned course of treatment—with all of the intricacies—might not be effectively and accurately communicated. An issue with any foreign language is the possibility there is an absence of a word or words, commonplace in English, for symptoms or findings. In Somali, there is a word for “swelling,” barar, but not “joint swelling.” There is no word for “stiffness,” but a direct translation would be dhaqaaq xumo (lack of movement). In northern Somalia, a translation might be sidii ay bir iku taagantahay (“as if there is a metal in my body”) (6).
Other aspects of appointments with Somali patients, be it the initial encounter or subsequent follow-up, are affected by not having the patient undress, respecting religious modesty. Though a reasonably thorough joint examination can be performed with a clothed individual, one might miss organomegaly or lymphadenopathy associated with some other illness, including neoplasm masquerading as inflammatory arthritis. Furthermore, one might miss finding an important underlying or associated illness by not obtaining optimal auscultation of the heart or lungs. By not asking about alcohol intake, does the physician assume the patient doesn’t drink because he or she is Muslim and does not want to offend? Perhaps the only reason that smoking is asked about is because of Meaningful Use requirements. Notably, smoking is an important risk factor for the development of RA, making it an essential question for the history (7,8).
Based on statistical expectations of RA, one would expect approximately 15 men to be a part of our database, yet there was only one. Again, the size of our group precludes making only speculative the observation that this reflects a stark difference in gender prevalence. Unique social expectations could raise the notion that Somali men may not raise concerns as much as women. One of the authors (Mahamed) suggests that this is not the case. Perhaps men are not inclined to take time from work, something difficult to compare in our study given that only five women worked outside of their home. One might also wonder if deeper reasons—financial issues, image, self-esteem, or social expression of masculinity—associated with the stress and mental health status of immigrants contributes to the observed disparity (9).
It has previously been reported (10) that follow-up in the Somali population is suboptimal, however, that report dealt with 100 patients initially presenting to an HCMC clinic, rather than a cohort with a specific diagnosis. We did not find kept appointments to be an issue with a low rate of cancellations or no-shows, in spite of our suburban location and challenges of transportation, especially in winter months. This could relate to the establishment of a specific and serious diagnosis and subsequent treatment with medications requiring regular monitoring. We do not have statistics for a comparable Caucasian population. We would like to believe that patients, their families, and their support networks understood the necessity and importance of discussions and responded accordingly, and that potential barriers were being recognized and addressed.
Physicians should initiate a necessary dialogue and not be afraid to ask difficult questions. Physicians also need to understand and appreciate cultural beliefs about disease (11,12,13) and factor them into caregiving. Community education about RA as well as other diseases can provide earlier access to appropriate medical attention. Physician education, perception, and self-reflection are essential to solving serious problems with healthcare delivery and a step toward negating implicit bias (14). Different patients have different needs, a notion that must not be forgotten with any ethnic minority, or even the majority population, and is a standard for solidifying important interpersonal relationships.
This is not a rigorously controlled medical study, but it is real life. Respect for the Somali culture and Islam appears to be at the heart of observations in our study. Social factors have the potential to play a dominant role in both the evaluation and decision-making process for physicians. Whether implicit bias or other simple, uncovered obstacles affect outcomes remains to be determined. By extension, we raise the question of how barriers might affect medical care of other illnesses in Somali or in other unique minorities.
Our results and observations raise many questions and, perhaps, provide the start of a pathway for providing improved care for any ethnic group.
1. With Somali patients, are we guilty of implicit bias and hampered in our diagnoses by the consequences of not having a patient undress or by not asking about alcohol or illegal drugs?
2. How do we integrate essentials of the health history as they might apply to different cultures? Is pain the same for a Somali woman as it is for a Caucasian man? What does loss of appetite mean to a Somali, someone who may have gone days without eating at various times during migration to the United States?
3. Does the presence of an interpreter foster implicit bias in addition to creating a barrier to communication? How do we better educate interpreters and family members as to the words we us in questions and explanations?
4. How do we better educate caregivers in terms of delivery of medical care to unique populations, as well as obtain better insights into the life of a refugee and the transformation that occurs after relocation?
5. How do we involve communities to raise awareness of various diseases and better facilitate access to care?
6. As scientists, how do we best assess characteristics of unique cohorts? Is the lack of smoking in our group or the younger diagnosis of our seropositive patients germane and deserving of further investigation?
7. How do we, as caregivers, reconcile implicit bias with our concern for showing respect? If not being respectful, are we guilty of explicit bias?
Paul Waytz, MD, is xxxxxxxxx. Andrew Forsberg, BA, is xxxxxx. Abdi Mahamed, MBA, is xxxxxxx.
1. Etiology and pathogenesis of rheumatoid arthritis. In Firestein GS, Kelley WN, eds. Kelley’s Textbook of Rheumatology. 8th ed. Philadelphia PA: Saunders/Elsevier: 2009: 1035-1086
2. New Census Date: Minnesota Somali population grows. Williams C (AP). Minneapolis Star-Tribune. October 21, 2011.
3. Aletaha D, Neogi T, Silman AJ, et al. 2010 rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative [published correction appears in Ann Rheum Dis. 2010; 69(10): 1892]. Ann Rheum Dis. 2010;69(9): 1580-1588
4. Tehlirian C, Bathon J. Rheumatoid arthritis clinical and laboratory manifestation. In: Klippel JH, Stone JH, Crofford LJ, et al., eds. Primer on the Rheumatic Diseases. 13th ed. New York, NY: Springer; 2008: 114-121
5. https://www.rheumatoidarthritis.org/ra/facts-and-statistics Last edited August 3, 2016
6. Personal communication: Warfa N, MSc PhD, University of Essex. July 22/2017
7. Bang SY, Lee KH, Cho SK, et al. Smoking increases rheumatoid arthritis susceptibility in individuals carrying the HLA-DRB1 shared epitope, regardless of rheumatoid factor or anti-cyclic citrullinated peptide antibody status. Arthritis Rheum. 2010;62(2): 369-377
8. Karlson EW, Lee M, Cook, MR, et al. A retrospective cohort study of cigarette smoking and risk of rheumatoid arthritis in female health professionals. Arthritis Rheum. 1999;42(5): 910-917
9. Warfa N, Curtis S, Watters C, et al. Migration experiences, employment status and psychological distress among Somali immigrants: a mixed-method international study. BMC Public Health. 2012;12: 749-761
10. Miller E, Gertner E, Quirk R, McCarty M. Rheumatologic diagnoses, characteristics, and needs of Somali patients referred to a rheumatology clinic serving the Somali population [abstract]. Arthritis Rheum. 2015;67 (supplement 10)
11. Samanta A, Samanta J, Johnson M. Rheumatoid arthritis in minority ethnic groups: patterns of disease, clinical and sociocultural features among British South Asians. Diversity in Health and Social Care. 2005;2: 99-118
12. Samanta A, Shaffu S, Panchal S, et al. Tinkering at the edges or collaborative symbiosis? Ethnicity and rheumatology: a consensual review discussion. Diversity and Equality Health and Care. 2012;9: 209-217
13. Pavlish CL, Noor S, Brandt J. Somali immigrant women and the American health care system: discordant beliefs, divergent expectations, and silent worries. Soc Sci Med. 2010;71(2): 353-361
14. Chapman EN, Kaatz A, Carnes M. Physicians and implicit bias: how doctors may unwittingly perpetuate health care disparities. J Gen Intern Med. 2013;11: 1504-1510