Abstract
Objective Long diagnostic delay remains an unsolved problem in many autoimmune rheumatic diseases (ARDs). One of the major contributing factors is poor symptom appraisal and the resulting delays in help-seeking by patients themselves. We therefore aimed to understand the symptom appraisal and help-seeking experience among patients with ARDs in a multiethnic urban Asian population and to explore its influencing factors.
Methods Semistructured interviews with 33 patients with ARDs were audio recorded and transcribed verbatim. We coded the transcripts deductively using the reported 3 stages of symptom appraisal (detection, interpretation, and response) as the framework, and inductively for newly emerging themes and subthemes.
Results All 3 stages of the symptom appraisal and help-seeking journey (ie, symptom detection [by self and by others], symptom interpretation [causes, consequences, and required actions] and symptom response [no action, self-management, seeking help from nonhealthcare professionals, and seeking help from healthcare professionals]) were observed among patients. Interactions among these 3 stages were also observed: symptom interpretation was found to influence subsequent symptom detection, and the outcome of symptom response was found to influence both subsequent symptom detection and symptom interpretation. Various personal and socioenvironmental factors (eg, knowledge and cultural beliefs about the symptom) that influenced symptom appraisal and help-seeking were identified from the interviews.
Conclusion The symptom appraisal and help-seeking journey of patients with ARDs is an iterative process of detection, interpretation, and response, and is influenced by various personal and socioenvironmental factors. Addressing modifiable factors could shorten the symptom appraisal and help-seeking interval and improve patient outcomes.
- help-seeking behavior
- psoriatic arthritis
- rheumatoid arthritis
- spondyloarthritis
- systemic lupus erythematosus
- theoretical models
Long diagnostic delay, namely the time lapse between a patient’s first symptoms and a diagnosis by a rheumatologist, remains an unsolved problem in many autoimmune rheumatic diseases (ARDs) and leads to delayed treatment initiation and poor patient outcomes.1-7 For example, the average delay between symptom onset and treatment initiation among patients with rheumatoid arthritis (RA), the most common ARD, is 12 months,5 exceeding the 3-month window of opportunity associated with better outcomes.8,9 Diagnostic delay can be divided into 4 intervals: (1) between symptom onset and the first visit to a healthcare professional (HCP; also referred to as symptom appraisal and help-seeking interval); (2) between the first visit to a HCP (primary care physicians and nonrheumatology specialists) and rheumatology referral; (3) between rheumatology referral and rheumatology assessment; and (4) between rheumatology assessment and diagnosis or treatment initiation.10 In contrast to the other intervals that are primarily influenced by physician- and healthcare system–related factors, the symptom appraisal and help-seeking interval is related primarily to patients themselves. It is thus important to understand patients’ symptom appraisal and help-seeking journey to address the issue of prolonged diagnostic delays.
Many theories and models have been proposed to study the symptom appraisal and help-seeking journey.11-20 In a review of these theories and models by Whitaker et al, 3 common constructs or stages of the symptom appraisal and help-seeking journey were identified: detection of bodily changes (symptom detection), interpretation of bodily changes (symptom interpretation), and response to interpretation (symptom response).21 According to these theories and models, bodily changes that arise from precipitating events such as an illness compete with other information for processing,22 and depending on the quantity and quality of the competing information, bodily changes may or may not be detected by symptomatic individuals.12,13,16,17 After the bodily changes are detected, they will be interpreted and first attributed to situational factors (normalizing attribution), and then—only if the normalizing attribution is insufficient—attributed to psychological distress (psychological attribution) and physical illnesses (somatic attribution).23 Decisions are then made in response to symptom interpretation.
However, how these symptom appraisal theories and models could be employed to shorten the symptom appraisal and help-seeking interval of individuals with possible ARDs in the general population remains unclear. In a systematic review of approaches to improving symptom appraisal, we established a theoretical framework in which approaches were developed based on the main concepts of social cognitive theory to improve patients’ symptom appraisal and help-seeking behaviors.24 Our primary objective in this study was therefore to understand the symptom appraisal and help-seeking experience among patients with ARDs, building on this theoretical framework. Our secondary objective was to explore the factors influencing the symptom appraisal and help-seeking journey in order to provide insights for future studies on the development of possible approaches to shorten the symptom appraisal and help-seeking interval and improve patient outcomes.
METHODS
Study setting. This study was performed at the rheumatology specialist outpatient clinic (SOC) at Singapore General Hospital (SGH), the largest public tertiary referral center in Singapore. SGH is located in the central region of Singapore but receives patients from all over the country. The rheumatology SOC in SGH has 15 accredited rheumatologists and manages > 25,000 outpatient visits each year (unpublished hospital data).
Singapore is a multiethnic urban population comprising individuals of Chinese, Malay, and Indian descent, and these represent 3 major ethnic groups in Asia.25 Similar to many other countries, patients with ARDs in Singapore are managed primarily in public or private rheumatology SOCs. Patients attending rheumatology SOCs could be referred from public or private primary care clinics (which offer same-day walk-in healthcare services) and nonrheumatology SOCs, or are self-referred. However, only patients attending public rheumatology SOCs who are referred from public primary care clinics are subsidized for medical expenses by the government.26
Study design. This study is part of a larger qualitative study to understand and explore possible approaches to improve the symptom appraisal and help-seeking journey of patients with ARDs. The reporting of findings was guided by the Consolidated Criteria for Reporting Qualitative Research (COREQ) checklist (Supplementary Material 1, available with the online version of this article).27
Study team. This study was designed and conducted by a group of researchers familiar with qualitative methods. LX, a medical doctor by training and a PhD student who is also trained in and has conducted several qualitative studies, facilitated the semistructured interviews. She worked closely with SY, a public health researcher and behavioral scientist, and JT, a clinician scientist and health services researcher, in the development of the interview guide and data analysis and interpretation, with important input from AHLL, YYL, WF, TCL, and DRK, clinician researchers who are experienced in mixed methods research.
Study participants. Participants in this study were drawn from those enrolled in an earlier study that assessed diagnostic delay (median 11.5 months) among patients who were newly diagnosed with ARDs in our rheumatology SOC.10 Purposive sampling was adopted in the present study, whereby patients who had participated in the earlier study and had > 3 months of diagnostic delay were invited to participate.10 Potential participants were approached in person during their routine clinic visits or by telephone. No relationship was established with the participants prior to the commencement of this study.
Data collection. An interview guide was developed based on the McGill Illness Narrative Interview initial illness narrative, prototype narrative, and explanatory model narrative, an interview protocol that has been widely adopted for use in various health issues and conditions.28 Relevant questions were revised to be culturally appropriate and rearranged into 3 sections—symptom detection, symptom interpretation, and symptom response—corresponding to the 3 main constructs of symptom appraisal identified by Whitaker et al (Supplementary Material 2, available with the online version of this article).21
Semistructured interviews were conducted by LX in the participant’s preferred language (English or Chinese). This was done in batches (approximately 5-6 interviews in each batch) to allow for concurrent data analysis and iterative modifications to the interview guide. Interviews were continued until data saturation was reached. Field notes and memos were taken during and immediately after the interviews. There were no nonparticipants present during the interviews. None of the interviews were repeated.
All interviews were audio recorded. Audio recordings were transcribed verbatim and, if needed, translated into English for analysis by trained research staff. Transcripts were not returned to participants for verification.
Data analysis. The transcripts were coded by LX and SY using the NVivo software (Release 1.7.1). We adopted the symptom appraisal component of our recently developed theoretical framework, namely the 3 main constructs of symptom appraisal identified by Whitaker et al, as our overarching framework.21,24
We coded the transcripts deductively using the Ritchie and Spencer framework analysis,29 and inductively for newly emerging themes and subthemes.30 Discrepancies in coding were resolved and the codebook was revised through iterative discussions between the 2 coders (Supplementary Table 1, available with the online version of this article). We adopted a hybrid form of data saturation involving a priori thematic saturation based on the overarching framework and an inductive thematic saturation at the level of code categories and themes and subthemes. No deviant cases were identified. The findings were not provided to participants for feedback.
Ethical approval. Ethical approval of this study was obtained from the SingHealth Centralised Institutional Review Board (reference no. 2020/2278). Written informed consent was sought from all participants, and the context and objectives of the study were explained to all participants prior to conducting interviews.
RESULTS
Study participants. Of the 50 patients we approached, 33 (66%) consented and completed the interview (Table 1). Reasons for declining participation included lack of interest in research studies (n = 9), concerns about face-to-face interviews during the coronavirus disease 2019 pandemic and/or discomfort with interviews by video conferencing (n = 6), and inability to recall the details of their symptom appraisal and help-seeking journey (n = 2).
Characteristics of participants.
Approximately 30% of the participants had RA, 15% had psoriatic arthritis (PsA), 12% had spondyloarthritis (SpA), and 12% had Sjögren syndrome. The other ARDs each accounted for < 10% of the participants (Table 1). The participants had a median age of 52.0 (Q1-Q3 43.0-58.0) years, and were mostly female (76%), Chinese (79%), and English-speaking (82%), with > 10 years of education (76%) and ≥ 1 comorbidity (76%). The median diagnostic delay and duration from diagnosis to interview were 12.0 (Q1-Q3 5.0-43.0) months and 28.0 ( 21.0-33.0) months, respectively. The interviews were conducted face to face in the rheumatology SOC (79%) or remotely by Zoom (21%). They lasted a median of 37.0 (30.0-43.0) minutes.
Study findings. There are 2 key findings from this study. First, all 3 main stages of the symptom appraisal and help-seeking journey (symptom detection, symptom interpretation and symptom response) were evident in patients with ARDs (Figure). In addition, interactions among these 3 stages were also observed. Second, various personal and socioenvironmental factors influencing the symptom appraisal and help-seeking journey were identified from the interviews. Key themes, subthemes, and illustrative quotes are presented in Table 2 and Table 3.
Interactions among the 3 stages of symptom appraisal and help-seeking, and influencing factors. Adapted from the synthesized model of symptom appraisal by Whitaker et al.21 The rectangles represent the 3 stages of the symptom appraisal and help-seeking journey. The rounded rectangles represent the 2 categories of factors influencing this journey. The double arrow represents the interactive relationship between personal and socioenvironmental factors.
Themes and illustrative quotes for the symptom appraisal and help-seeking journey.
Themes and illustrative quotes for factors influencing the symptom appraisal and help-seeking journey.
Symptom appraisal and help-seeking journey.
∙ Symptom detection. All participants shared that they had ≥ 1 symptom or bodily change relevant to their ARD before they sought help from HCPs. The majority of the participants reported that their symptoms were first detected by themselves, whereas the minority were not aware of their symptoms until they were told by other people, including their family members, friends, colleagues, and HCPs (when these participants were seeking medical attention for other preexisting conditions). Compared to symptoms that were first detected by participants themselves, symptoms that were first noticed by other people were of lesser intensity and/or caused lesser disruption to participants’ daily life (eg, self-care; Table 2, quotes 1-2).
∙ Symptom interpretation. Following symptom detection, most participants reported they had tried to figure out the possible causes and consequences of their symptoms, and what actions were required.
The majority of participants shared that they had misattributed their symptoms to various causes, including environmental (eg, coldness in an air-conditioned room); psychological (eg, stress); or other issues such as aging, diet (eg, nonnutritious food intake), overexertion, or preexisting conditions (eg, injuries; Table 2, quote 3). Only a few participants suspected that they might have ARDs due to their positive family history (Table 2, quote 4). The remaining participants shared that they could not pinpoint any reasons for their symptoms as there were no precipitating events prior to their symptom onset (Table 2, quote 5).
Many participants reported that they had thought about the possible consequences of their symptoms, based primarily on the characteristics and self-perceived causes of symptoms. Some participants reported that they were terrified by the great intensity of their symptoms (eg, severe pain) and/or the significant disruption to their daily life (eg, inability to get up) and thought that their symptoms could lead to serious consequences (Table 2, quote 6). In contrast, other participants had less severe and disruptive symptoms and believed that their symptoms were not a big deal and/or “would just go away” (Table 2, quote 7).
Based on the self-perceived causes and/or consequences of symptoms, many participants made decisions on whether actions were required and, if so, which actions were required (eg, managing the symptoms on their own vs seeking help from HCPs or non-HCPs) in response to their symptoms (Table 2, quotes 8-9).
∙ Symptom response. Symptom responses shared by the participants can be grouped into 4 categories: no action (ignorance or observation), self-management, seeking help from non-HCPs (laypersons), and seeking help from HCPs.
The typical symptom responses shared by many participants was that they initially decided to take no action but rather opted to ignore or observe their symptoms (Table 2, quote 10). When the symptoms deteriorated (eg, became more consistent), some of these participants tried to self-manage and/or seek help from non-HCPs before they sought help from HCPs, whereas others decided to seek medical attention directly. Notably, only a small proportion of the participants reported that they decided to seek help from HCPs immediately after the detection of their symptoms.
Self-management strategies shared by the participants included changing their diet, exercising, massaging, applying creams or ointments, using hot or cold compresses and medicated plasters, and taking over-the-counter medications (Table 2, quote 11). Non-HCPs that had been consulted by the participants included their family members, friends, colleagues, and other people with the same or similar symptoms (including strangers online), and even unlicensed therapists (Table 2, quote 12). Some participants reported that their self-management or the management suggested or offered by non-HCPs relieved their symptoms to different extents temporarily, whereas others reported no effects at all.
Nearly all participants reported that they had consulted an HCP in public or private primary care clinics before they were referred to the rheumatology SOC. Some participants reported they were referred on their first visit to HCPs, whereas others visited the same HCP several times or visited various HCPs before they were referred to rheumatologists (Table 2, quotes 13-14).
∙ Interactions in the symptom appraisal and help-seeking journey. An important common theme that ran through the interviews was the iterative nature of the symptom appraisal and help-seeking journey: symptom interpretation influences subsequent symptom detection, and the outcome of symptom response influences subsequent symptom detection as well as symptom interpretation. Some participants reported that they became more proactive in checking their body to see if the same symptoms would reoccur and/or to look out for new symptoms or bodily changes if they perceived their symptoms to be caused by something serious, or after they became aware of the significance of their symptoms (Table 2, quote 15). The temporary alleviation of symptoms reinforced the validity of participants’ symptom interpretation, whereas no change or deterioration of symptoms would prompt the participants to recognize other characteristics of their current or new symptoms, and to realize that their symptoms could be due to other causes (Table 2, quote 16). This change of symptom interpretation would, in turn, change participants’ subsequent symptom response.
Factors influencing the symptom appraisal and help-seeking journey. Because the 3 stages of the symptom appraisal and help-seeking journey interact with one another, factors influencing 1 stage of the journey may also influence the other 2 stages. Therefore, we grouped the factors identified from the interviews into 2 interacting categories: personal factors and socioenvironmental factors.
Personal factors. The identified personal factors included symptom characteristics, knowledge of the symptom, and demographic characteristics. Symptoms of greater intensity and/or those causing greater disruption to daily life were more likely to be detected and deemed to have serious consequences by the participants (Table 2, quote 1 and quote 6). Participants with such symptoms more frequently reported seeking help from HCPs, whereas those with less serious symptoms more frequently reported no action, self-management, or seeking help from non-HCPs (Table 3, quotes 1-2).
Insufficient knowledge of a symptom (eg, possible causes) resulted in incorrect symptom interpretation and was deemed a key barrier for promptly seeking help from HCPs by many participants. The majority of these participants shared that they were not aware of the possibility of having severe conditions like ARDs and that they did not have any knowledge about ARDs before they were diagnosed with one (Table 3, quote 3). A few participants reported that although they were aware of the possibility of having severe conditions, they chose to deny it instead because they were worried that their conditions could not be easily managed (Table 3, quote 4).
Demographic characteristics such as old age and lower socioeconomic status were also reported to influence participants’ symptom interpretation and response. Old age was believed to be the main cause of symptoms by many participants who, as a result, thought it was unnecessary to seek help from HCPs (Table 2, quote 3). Some participants reported that the anticipated high medical expenses could be an important barrier for the less well-off patients to seek medical attention (Table 3, quote 5).
Socioenvironmental factors. The identified socioenvironmental factors included social relations, cultural beliefs about the symptoms, and the healthcare system. The social role (eg, housewife) was deemed to contribute to the symptoms by some participants (Table 3, quote 6). Family members and significant others of many participants were reported to assist in their symptom detection (Table 2, quote 2) and influence their symptom interpretation (Table 3, quote 6) and response (eg, seeking help; Table 3, quote 7).
Cultural beliefs about symptoms played an important role in the participants’ symptom interpretation and response, especially when the knowledge of the symptom was insufficient. In a culture where certain symptoms are considered normal (eg, hands and toes being sensitive to cold in females) or could not be noticed by other people (eg, certain parts of a female’s body are covered due to religious beliefs), participants with these symptoms did not seem to seek help from HCPs until their symptoms progressed and/or started to significantly disrupt their daily life (Table 3, quotes 8-9).
Due to the healthcare system and financing schemes in Singapore, the vast majority of participants reported that they had sought help from general practitioners (GPs) in public primary care clinics before they were referred to the public rheumatology SOC (Table 3, quote 10). However, depending on which GP a participant visited, there may have been a misdiagnosis and/or delay before they were referred for further assessment (Table 2, quotes 13-14).
DISCUSSION
In this study, we explored the symptom appraisal and help-seeking experience and its influencing factors among patients with ARDs in a multiethnic urban Asian population using the common constructs of symptom appraisal theories and models as a framework. We found that symptom detection, symptom interpretation, and symptom response were 3 essential and interacting stages in the symptom appraisal and help-seeking journey, and that this journey was influenced by the interplay of personal and socioenvironmental factors (eg, knowledge and cultural beliefs about symptoms). Our findings add to the literature by demonstrating the validity of symptom appraisal theories and models and identifying the factors influencing symptom appraisal and help-seeking behaviors in the context of ARDs. Addressing the factors identified in our study could potentially shorten the symptom appraisal and help-seeking interval among individuals with possible ARDs in the general population and improve patient outcomes.
Symptom misinterpretation was found to be the main driver of long diagnostic delays among our participants with ARDs, as symptom response is dependent primarily on patients’ symptom interpretation. Owing to the incorrect symptom interpretation seen in most participants, observation, self-management, and seeking help from non-HCPs were the most common symptom responses and have been reported frequently among patients with various ARDs in the literature.31-38 This highlights the importance of improving symptom interpretation, which in turn would promote patients to seek help from HCPs earlier. Further, in light of the iterative nature of the symptom appraisal and help-seeking journey, addressing factors influencing the symptom interpretation stage could not only correct the misinterpretation of symptoms but also prompt individuals to proactively look out for possible symptoms (symptom detection) and promote patients to seek help earlier on from HCPs (symptom response).
Insufficient knowledge and cultural misbeliefs about symptoms are key barriers to seeking earlier help from HCPs among our participants with ARDs. Because insufficient knowledge existed in the majority of the participants prior to their diagnosis, cultural beliefs about symptoms played an important role in the decision-making process by shaping the perception of the causes and consequences of and actions required for a given symptom. Importantly, these perceptions and consequent actions often contributed to delays in seeking medical attention. For symptoms that were believed to be the consequences of minor issues that could potentially be solved by oneself, seeking help from HCPs was often not the first action of choice. For example, cold sensitivity was believed to be the consequence of not eating enough red food (in color, such as red dates) and meat among the Chinese people. Thus, individuals with cold sensitivity due to Raynaud phenomenon might try to solve this by eating more red food and meat first. For symptoms that do not have an explanation in the culture, one’s symptom response could also be shaped by the perceived consequences of the symptom, based on their cultural practices. For example, hair loss among Malay women was believed to require no action as their hair could not be seen by other people (except their husbands and other family members) due to the use of a head covering for religious reasons. Approaches to addressing the insufficient knowledge of symptoms may also help dispel cultural misbeliefs and prompt symptomatic individuals and their significant others to recognize possible symptoms of severe conditions like ARDs, which allows for a supportive environment for early help-seeking. One possible approach to address the insufficient knowledge of the symptom among symptomatic individuals is to raise the awareness of ARDs through public education. Though challenging given the rarity of ARDs in the general population, public awareness efforts have shown promising results in facilitating early help-seeking and diagnosis.39,40
Apart from the knowledge and cultural beliefs about symptoms, social role, and influence from family members and significant others, other factors influencing the symptom appraisal and help-seeking journey—namely demographic characteristics and the healthcare system—are not readily modifiable; however, efforts could be made to address them indirectly. For example, lowering medical expenses and/or increasing subsidies for medical expenses might encourage symptomatic individuals with lower socioeconomic status to seek early medical attention.
The main strengths of this study are that it illustrates the 3 essential and interacting stages of the symptom appraisal and help-seeking journey of patients with ARDs and that it identifies factors influencing this journey. Our findings can provide useful insights into the development of possible approaches to shortening the symptom appraisal and help-seeking interval among symptomatic individuals in the general population. The main limitation of this study is the potential recall bias due to the relatively long time lapse between diagnosis and interviews. However, participants were given sufficient time (a few days to weeks) upon consenting to recall their symptom appraisal and help-seeking journeys prior to the interview, which would help to minimize the recall bias. Another limitation is that all participants in our study were recruited from a public rheumatology SOC. Some of the factors that had influenced the symptom appraisal and help-seeking journey of these participants might not be applicable to patients seeking treatment in private rheumatology SOCs, where medical expenses are not subsidized by the government. For example, patients seen in private rheumatology SOCs are often in higher socioeconomic classes and may not have concerns about the high medical expenses.
The symptom appraisal and help-seeking journey of patients with ARDs is an iterative process of detection, interpretation, and response and is influenced by various personal and socioenvironmental factors. Addressing modifiable factors could shorten the symptom appraisal and help-seeking interval and improve patient outcomes.
ACKNOWLEDGMENT
The authors thank Ms. Hui-Ling Tan for her assistance in participant recruitment and transcription, and all patients who participated in this study.
Footnotes
This study was funded by the SingHealth Duke-NUS Academic Medicine Research Grant (grant reference no. AM/HSR015/2021) and Goh Cheng Liang Rheumatology Advancing Research and Innovation with Synergistic Expertise (ARISE) Program Fund. The study funders had no role in the design of the study and collection, analysis, and interpretation of data and in the writing of this manuscript.
The authors declare no conflicts of interest relevant to this article.
- Accepted for publication July 20, 2023.
- Copyright © 2023 by the Journal of Rheumatology
