Healthcare-related social determinants of health are associated with pain severity among adults with comorbid heart failure and chronic pain

Abstract

Introduction

Chronic pain worsens symptom burden and quality of life in individuals with heart failure (HF). Healthcare-related social determinants of health (SDOH) - including access (e.g., stability of coverage), affordability (e.g., medication cost-saving activities), and utilization (e.g., number of provider visits) may influence pain severity. This study aimed to describe these SDOH and examine their relationship with pain severity.

Methods

A cross-sectional analysis was conducted using data from 4,026 adults with HF and chronic pain enrolled in the All of Us Research Program. Pain severity over 7 days (0 to 10 scale) and healthcare-related SDOH were assessed using descriptive statistics and linear regression.

Results

Participants averaged 69.90 years old; 58.15% were female and 68.11% White. Although 98.48% had insurance, barriers remained: 21.82% reported transportation issues, 22.45% could not afford dental care, and 34.88% requested lower-cost prescriptions. Participants reporting more barriers to healthcare and more unaffordable healthcare services experienced significantly higher 7-day pain severity. Other variables associated with higher 7-day pain severity were age, race, sex, comorbidity, depression and fatigue.

Conclusions

Despite high insurance coverage, healthcare barriers were common and associated with increased 7-day pain severity. Future research should explore access to pain-specific services with more comprehensive pain measures.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12913-026-15009-9.

Introduction

Heart failure (HF) affects more than 6 million people in the United States [1], and prevalence is expected to increase to more than 10 million people by 2050 [2]. Individuals with HF often experience a debilitating symptom profile which can include pain, dyspnea, depression, and sleep disturbances [3–5]. This symptom burden decreases physical function [3], quality of life [6, 7], and self-management capacity [8]. Crucially, comorbid pain is one common symptom in HF (23.00% to 85.00%) [9], that is often reported more frequently than classic symptoms including dyspnea and fatigue [6]. Although diagnostic and treatment guidelines for chest pain are well established in cardiovascular disease [10], recommendations for other types of pain in HF are absent [11, 12]. In addition, many adults with HF report little or no relief from their existing pain treatment regimens [5].

There is a critical lack of empirical research exploring healthcare-related social determinants of health (SDOH), particularly healthcare access and quality. This is because of the noted access disparities to guideline-directed therapies among vulnerable populations with HF [13], which may also extend to management of pain. Chronic pain in particular, which lasts for months to years [14], represents a significant problem that requires long-term management coordination with a healthcare team, which may be impacted by SDOH. The SDOH are defined as the conditions in which individuals live and interact [15], consisting of economic, education, healthcare, neighborhood, and social domains [15, 16]. The healthcare domain pertains to healthcare access, affordability and utilization, Healthcare access is the ability to obtain healthcare services when needed to improve outcomes [17]. Optimal healthcare access involves the presence of applicable healthcare systems, the means to enter into these systems, and the personnel within these systems who can provide the needed services. Affordability is another complex and subjective determinant of individual health, where value of healthcare goods and/or services are weighed against the available resources to obtain them [18]. One final approach to quantifying the downstream impact of healthcare-related SDOH is to examine healthcare utilization, which represents a quantification of the healthcare services used by individuals [19].

While a relationship between SDOH and pain in HF has been proposed in the theoretical literature [20, 21], these relationships are understudied empirically. Authors of a recent systematic review identified almost 60 studies of SDOH in HF, yet the only healthcare access-related SDOH that were included were insurance status (n = 27), and access to a cardiologist (n = 1) [15]. Other self-reported barriers such as distance, perceived bias, financial limitations, or caregiver burden were not included in the studies reviewed, and none of these studies measured pain to our knowledge [9, 22]. However, a growing body of research has recognized that economic and environmental conditions (e.g., housing, employment, healthcare access) impact pain and contribute to health inequities in non-HF populations [23, 24].

There is a need for studies examining a comprehensive set of healthcare-related SDOH among adults with comorbid HF and chronic pain to improve understanding of pain impact. The aims were: (1) describe healthcare access, affordability and utilization among adults with HF and chronic pain; and (2) examine the relationship between healthcare access and affordability with pain severity among adults with HF and chronic pain.

Methods

Design and data sources

The study design was retrospective and cross-sectional. The data were obtained from the All of Us Research Program (AOURP), a national registry consisting of over 620,000 participants at the time of the current study. The AOURP has to date enrolled a diverse sample that is representative of the demographic makeup of the United States [25]. Over 80.0% of participants in the AOURP are from populations historically underrepresented in research, including race, access to healthcare, and geographical location [26]. All participants provided informed consent before enrolling in the AOURP. The National Institutes of Health Institutional Review Board approved the AOURP, and the current study was exempted by the Indiana University Institutional Review Board (22864). The study adhered to the principles outlined in the Declaration of Helsinki. Data sources from the AOURP utilized in this study included surveys and medical records obtained from consented participants.

Sample

Access to the deidentified AOURP Registered Dataset v8 was obtained within the Research Workbench. Over 195,000 participants had the necessary surveys (i.e., “The Basics”, “Overall Health”, and “Healthcare Access/Utilization”) and medical records available for the study. The inclusion criteria consisted of (1) a diagnosis of chronic pain; and (2) a diagnosis of HF. Diagnoses of chronic pain and HF were both established using International Classification of Diseases (ICD) 9 and 10 codes available in medical records [27]. Only ICD codes entered in the medical record before the participant completed the AOURP surveys were included. The use of ICD codes as an approach to classifying presence of HF, pain, and multimorbidity in large samples has been previously published [28–30]. The exclusion criteria consisted of (1) no historical medical records available; and (2) missing data for variables used in the analysis. After applying the inclusion and exclusion criteria, the final sample size was 4,026 participants. A flow diagram detailing the process for creating the dataset used in the current study is provided in Fig. 1.

Fig. 1.

Flow diagram of participants included in the current study. Note. AOURP: All of Us Research Program; HF: heart failure; SDOH: social determinants of health

Measures

Pain severity was assessed using a single-item question - “In the past 7 days, how would you rate your pain on average?” - which was scored on a 0 to 10 scale [31]. Higher scores indicated more severe pain. A value of 0 represented no pain, while a value between 1 and 3 represented mild pain, 4 to 7 represented moderate pain, and 8 to 10 represented severe pain. This pain item originates from the Patient-Reported Outcomes Measurement Information System Global Health measures, with previously published evidence of validity and reliability [32]. Given that this item only assess pain within a 7 day period, it was not assumed to be necessarily related to a chronic pain condition despite all participants having such a diagnosis.

To address the aims, instruments measuring healthcare access, affordability, and utilization were leveraged in the AOURP to operationalize healthcare-related SDOH (see Supplemental Table 1). These were derived from either the National Health Interview Survey [33], or the American Community Survey [34]. Healthcare access consisted of four instruments from the AOURP: (1) insurance status; (2) insurance acceptance; (3) insurance coverage stability; and (4) number of self-reported access barriers to healthcare. Insurance status was assessed through one question asking if the participant had insurance (yes, no). Insurance acceptance was assessed through one question asking if the participant was told their health insurance was not accepted in the past 12 months (yes, no). Coverage stability was assessed through one question asking how their perceived healthcare coverage compared to one year ago (worse, about the same, better). Self-reported access barriers reported by participants consisted of 9 possible options (e.g., lack of transportation, distance), and were transformed into a single continuous variable ranking from 0 to 9. A higher score indicated that more barriers were present.

Healthcare affordability consisted of two instruments: (1) the number of healthcare services that were viewed as unaffordable, consisting of 8 possible options (e.g., prescriptions, mental health counseling); and (2) the number of medication cost-saving activities undertaken, consisting of 6 possible options (e.g., skipped medication, took less medication). As with barriers, both these variables were each transformed into continuous variables (range 0 to 8 for unaffordable healthcare services, 0 to 6 for medication cost-saving activities). A higher score on these two instruments indicated that more services were reported as unaffordable, or more medication cost-saving activities were undertaken by the participant, respectively.

Healthcare utilization consisted of three instruments: (1) the number of general doctor visits; (2) the number of prescribing advanced practice provider visits (i.e., nurse practitioner, physician assistant, or midwife); and (3) the number of specialist visits (i.e., physicians focusing on a particular medical condition excluding obstetricians/gynecologists, ophthalmologists, or psychiatrists). The number of provider visits was measured over a 12-month period. The original data structure for these instruments consisted of categories representing a range of visits, rather than a continuous variable (e.g., “2 to 3 visits”). Because of this, four categories were recoded from the available data to represent healthcare utilization. These included low (0 to 1 visit), average (2 to 3 visits), high (4 to 7 visits), and very high (8 + visits). These categories were based on previous estimates of the number of visits per person per year for primary care providers (1.61) [35], nurse practitioners (1.70) [36], and specialists (1.59) [35].

Other measures included due to previous evidence supporting a relationship to pain in HF were age [22, 37–39], natal sex [40, 41], self-reported race [42], fatigue (average over 7 days, 1 to 5 scale with a higher score indicating lower fatigue) [3, 4, 39], depression diagnosis (obtained from the medical record) [3, 37–39, 43], and number of comorbidities (i.e., none, one disease, two diseases, three or more diseases). Comorbidities included were chronic kidney disease [44], chronic obstructive pulmonary disease [45], coronary artery disease [22], diabetes, myocardial infarction [22, 41], peripheral vascular disease [46], and stroke [46], and were obtained from the medical record.

Statistical analyses

Individual items were treated as missing if the participant indicated “don’t know”, “skip”, or “prefer not to answer”. For variables used in the regression analyses, a complete case approach was applied as these variables had 5.00% or less total missing data. The utilization data were analyzed descriptively because of > 5.00% missing data for the number of advanced practice provider visits (14.22%), and number of specialist provider visits (22.19%). To identify uneven missing data in vulnerable populations, the percentage of missing data by natal sex and by race was examined. The largest differential in missing data for pain severity was between participants reporting Asian race vs. reporting more than one race (9.62% vs. 2.95%, respectively) (see Supplemental Table 2). Univariate analyses comparing sociodemographic variables and insurance status between missing data and non-missing data groups were conducted, consisting of independent samples t-tests for continuous variables and Chi [2] for categorical variables, to determine whether excluded participants differed systematically.

All analyses were conducted using SAS in the AOURP Research Workbench. A statistical significance level was denoted as p < .05. Characteristics of the sample were calculated using descriptive statistics, including means, proportions and standard deviations. For aim 1, descriptive statistics were used to profile the healthcare-related SDOH in the sample. The number of visits were tabulated by 7-day average pain severity categories (none, mild, moderate, and severe). For aim 2, multiple linear regression was used to examine the contribution of selected healthcare-related SDOH on 7-day average pain severity while controlling for age, sex, race, number of comorbidities, depression, and fatigue. Multicollinearity was assessed using variance inflation factors, correlation of the predictors, and examination of plotted residuals and the standard errors of the estimated effects.

Results

Demographic and clinical characteristics of the sample are provided in Table 1. The sample primarily consisted of older adults. The breakdown of natal sex consisted of 58.15% female, 41.85% male. Most of the sample reported White race (68.11%). More than 86.00% of the sample had at least one comorbid condition. The average fatigue scores were 3.30 out of 5 (SD 0.95) and average 7-day average pain severity scores were 4.61 out of 10 (SD 2.88). The percent of participants who reported either moderate or severe pain (i.e., reported a 7-day average pain severity score ≥ 4 out of 10), was 61.17%. Based on univariate analyses, age and fatigue were statistically different between participants with vs. without complete data. However, the actual differences were minimal: a 0.76% difference in age (69.90 years with complete data vs. 70.43 years with incomplete data), and a 1.9% difference in fatigue (3.30/5 with complete data vs. 3.36/5 with incomplete data) (See Supplemental Table 3).

Table 1.

Demographic and clinical characteristics (N = 4,026)

Variable	Mean ± SD	No, %
Age in years,	69.90 ± 6.96
Natal sex
Female		2341 (58.15%)
Male		1685 (41.85%)
Race
Asian		32 (0.79%)
Black		628 (15.60%)
More than one race		198 (4.92%)
Other		426 (10.58%)
White		2742 (68.11%)
Comorbidity
None		558 (13.86%)
One disease		1126 (27.97%)
Two diseases		957 (23.77%)
Three or more diseases		1385 (34.40%)
Depression		1744 (43.32%)
Fatigue	3.30 ± 0.95
Pain severity (7-day period)	4.63 ± 2.89
Pain groups
No pain (7-day severity 0)		360 (8.94%)
Mild (7-day severity 1–3)		1195 (29.68%)
Moderate (7-day severity 4–6)		1710 (42.27%)
Severe (7-day severity 7–10)		761 (18.90%)

Notes – SD: standard deviation

Healthcare access, affordability and utilization variables in the sample are described in Table 2. Over 98.00% of participants had health insurance in the previous year, and 10.71% had had their health insurance denied during the previous year. More than 77.0% of the sample reported that their coverage status felt about the same compared to a year ago. Over 62.00% of the sample had seen a general doctor 4 times or more in the past 12 months. In addition, 42.47% and 43.57% of the sample saw an advanced practice provider or specialist 2–3 times in the past 12 months, respectively.

Table 2.

Healthcare access, affordability and utilization variables among adults with heart failure and chronic pain (N = 4,026)

Variable	Mean ± SD	No., %
Healthcare access
Insurance status
No		61 (1.52%)
Yes		3965 (98.48%)
Insurance acceptance
No		3595 (89.29%)
Yes		431 (10.71%)
Coverage stability
Better		573 (14.23%)
About the same		3103 (77.07%)
Worse		350 (8.69%)
Number of self-reported access barriers	0.69 ± 1.18
Healthcare affordability
Number of unaffordable healthcare services	0.89 ± 1.64
Number of medication cost-saving activities	0.70 ± 1.27
Healthcare utilization
General doctor visits per 12 months (n = 3,899)
0–1		435 (11.16%)
2–3		1046 (26.83%)
4–7		1274 (32.68%)
8+		1144 (29.34%)
Advance practice provider visits per 12 months (n = 3,626)
0–1		1,530 (42.2%)
2–3		1,540 (42.47%)
4–7		431 (11.89%)
8+		125 (3.45%)
Specialist visits per 12 months (n = 3,608)
0–1		1,050 (29.10%)
2–3		1,572 (43.57%)
4–7		782 (21.67%)
8+		204 (5.65%)

Pie charts depicting the different types of self-reported access barriers, unaffordable healthcare services, and medication cost-saving activities are presented in Fig. 2. The most common self-reported barriers that delayed care were no transportation (21.82%) and having to pay out of pocket (19.54%). The most common unaffordable healthcare services were dental care (22.45%), and prescription medicines (20.80%). The most common medication cost-saving activities undertaken by participants were asking for lower cost medications (34.88%) and delaying filling a prescription (21.90%).

Fig. 2.

Self-reported access barriers, unaffordable healthcare services, and medication cost-saving activities among adults with diagnoses of HF and chronic pain (N = 4,026). Note. Participants could indicate more than one category

A comparison of general doctor, advanced practice provider, and specialist visits within a 12-month period organized by 7-day average pain severity groups (none, mild, moderate, or severe) is presented in Fig. 3. Of participants reporting no pain, 49.14% saw a general doctor four times or more in the previous 12 months. In contrast, 72.75% of participants experiencing severe pain on average saw a general doctor at the same frequency level. By comparison, 24.78% of participants with severe pain saw an advanced practice provider four times or more, and 30.69% of participants saw a specialist four times or more.

Fig. 3.

Healthcare utilization within the past 12 months by 7-day average pain severity groups among adults with diagnoses of heart failure and chronic pain

The linear regression examining the effect of healthcare access and affordability variables on 7-day average pain severity is presented in Table 3. Participants who perceived a positive shift in insurance coverage in the past year reported higher 7-day average pain severity scores compared to those who had reported no change. In addition, participants who reported a higher number of access barriers and a higher number of unaffordable healthcare services reported higher 7-day average pain severity. Participants who were younger, female sex, self-reported Black, other, or more than one race, had higher comorbidity, had a positive depression diagnosis, and reported more fatigue all reported significantly higher 7-day average pain severity scores. The adjusted R² of the final model was 26.83%.

Table 3.

Linear regression of healthcare access and affordability variables on 7-day average pain severity (N = 4,026)

Variable	B ± Std. Error	t	p-value
Healthcare access
Insurance status (ref: yes)	-0.046 ± 0.322	-0.14	0.885
Insurance acceptance (ref: yes)	-0.109 ± 0.131	-0.83	0.405
Coverage shift (ref: about the same)
Better	0.484 ± 0.114	4.26	< 0.001
Worse	0.180 ± 0.145	1.24	0.215
Self-report access barriers	0.220 ± 0.042	5.24	< 0.001
Healthcare affordability
Unaffordable services	0.137 ± 0.032	4.22	< 0.001
Money saving activities	-0.090 ± 0.039	-2.34	0.019
Covariates
Age	-0.025 ± 0.006	-4.37	< 0.001
Sex (ref: Male)
Female	0.679 ± 0.083	8.19	< 0.001
Race (ref: White)
Asian	0.010 ± 0.440	0.02	0.982
Black	1.06 ± 0.112	9.45	< 0.001
More than one race	0.943 ± 0.183	5.15	< 0.001
Other	0.841 ± 0.131	6.45	< 0.001
Comorbidity (ref: None)
One disease	0.363 ± 0.129	2.80	0.005
Two diseases	0.403 ± 0.133	3.03	0.003
Three or more diseases	0.798 ± 0.127	6.28	< 0.001
Depression	0.167 ± 0.081	2.06	0.040
Fatigue	-1.063 ± 0.044	-24.31	< 0.001

Discussion

This study contributes towards recent calls for research examining SDOH among individuals with HF [13]. What this study adds is a comprehensive examination of healthcare-related SDOH, including access, affordability, and utilization among adults with HF and chronic pain, which to our knowledge extends beyond results reported in previous studies [15]. In addition, the AOURP dataset contains a diverse sample with historically underrepresented groups which reduces the threat of selection bias. Given that healthcare access barriers play a role in symptom experiences among adults with HF generally [47], it is unsurprising that such access barriers were pronounced in a sample with HF and chronic pain. The results of this study carry important implications at both a provider and systems level, but additional research is needed to address the limitations of this secondary dataset.

The associations between affordability and 7-day average pain severity reveal that chronic pain in adults with HF is not just a clinical issue, but one deeply intertwined with systemic access barriers. Efforts to improve healthcare access based on the results could include reducing transportation and out-of-pocket cost barriers, including proactive financial counseling for medications. Utilization of telemedicine or remote care options may also improve pain management, but this has yet to be studied in HF. Transitioning to ICD-11 coding would better facilitate diagnostic specificity and accuracy for pain. A substantial body of research has highlighted the limitations of ICD-9 and ICD-10 codes in identifying chronic pain [48]. While ICD-11 codes offer greater diagnostic specificity and accuracy [49], they have not yet been widely implemented in US clinical practice [50], and as such are not available in AOURP. Once ICD-11 is broadly adopted, future studies can be designed to focus on distinct, easily identifiable pain phenotypes.

The findings present additional provider level implications. While primary care and outpatient HF clinics consist of the most accessible entry points, the heavy reliance on general doctors could indicate a potential misalignment of resources. This utilization pattern is likely exacerbated by the lack of recommendations for non-chest pain in HF from current clinical guidelines [11, 12]. Despite this gap, recommendations in other clinical populations are to include a multidisciplinary team including nurses and pain specialists, which reduces the burden on individual providers, and increases effectiveness of management efforts [51]. The lower utilization of advanced practice providers or specialists by adults with HF and chronic pain is particularly notable because these providers are often central to multidisciplinary care teams. Following future research, workflow changes may include a pain severity trigger which automatically refers patients to pain specialists within a multidisciplinary team following a presentation with moderate or severe pain. A related approach may be integrating a co-management visit between general and advanced practice providers, which has been shown to greatly improve non-pain outcomes [52]. However, given that the utilization measures available in AOURP are not specific to pain services, additional targeted research is needed to explore these potential relationships in more detail before interventions can be implemented. Ultimately, a combination of both systems level changes with provider engagement will be crucial to mitigate the negative effects of undertreated pain.

Defining the salient SDOH barriers in this population is crucial for successful implementation of new treatments. For example, the effectiveness of pain interventions that utilize outpatient centers, multidisciplinary health teams, or mobile or digital health applications, are lessened without addressing pertinent access barriers. While many clinical trials are designed to anticipate and account for these barriers (e.g., travel vouchers, provision of cell phones or tablets), these provisions are typically not maintained following the trial conclusion, necessitating a more sustained effort to ensure long term effectiveness. This is an important consideration given the number of travel, financial, and caregiver barriers observed in the sample. Given the high frequency of lower-cost medication requests, newly introduced medications for pain that are prohibitively expensive may be at a higher risk of either being delayed, discontinued, or not initiated at all. Ultimately, these data offer a unique perspective regarding optimal implementation of novel pain management interventions.

Several findings should be evaluated in subsequent research. An unexpected finding was that a higher number of medication cost-saving activities was significantly associated with lower pain severity scores. Originally, it was proposed that participants adopting more cost-saving activities would report increased 7-day pain severity because of an inability to acquire needed medications or therapies. It may instead be that these participants who were experiencing less pain were in fact able to perform these cost-saving activities more frequently, whereas those with higher pain could not afford the consequences of doing so. A supplemental model was conducted with medication cost-saving activities as the dependent variable and pain severity as the independent variable (all other variables retained), that produced the expected negative association between cost-saving activities and pain (estimate − 0.016, p = .013), which offers support for this explanation (Supplemental Table 4). However, future research is needed to better understand the complex decision-making processes underlying medication and other health-related cost-saving activities, which are likely highly individualized. Another finding that warrants additional study is that participants reporting a positive change in insurance coverage also reported higher 7-day average pain severity when compared to individuals reporting no change in coverage. The temporal direction of this relationship is unknown, but plausible explanations could include a sick-role or health-seeking effect, in which increasing chronic disease burden prompts individuals to obtain better coverage; survivorship, where those experiencing the most severe pain are more likely to have secured improved coverage by the time of the survey; or residual confounding related to disease severity. Participants who self-reported black, other, or more than one race reported higher 7-day average pain severity. A recent scientific statement highlighted the challenges accessing healthcare among racial and ethnic minority groups, which may also pertain to access to services for chronic pain [53]. As one of the largest and most diverse cohorts of individuals which focuses on previously underrepresented populations, the AOURP could be further leveraged to examine the degree of health disparities in HF and pain. However, it is also important to recognize that a single race variable not be able to fully capture the diverse experiences of pain among individuals who derive from a wide variety of cultural backgrounds [54]. Notably, the AOURP surveys contains additional variables that may have added context to the included race variable, such as medical gaslighting, neighborhood quality, or perceived bias. These could be included in a follow-up study that focuses primarily on the effects of discrimination and pain, rather than healthcare related SDOH [24]. Lastly, while comorbidity was significantly associated with 7-day average pain severity, additional follow up targeted studies should ascertain the unique contributions of each comorbidity. This is because certain comorbid conditions are linked with different phenotypes of pain (e.g., diabetic neuropathy in diabetes), and a combination of several pain presentations would complicate attempts at management.

Limitations

The primary limitation is that the available pain severity item was not phrased to pertain to chronic pain specifically. This impacts the validity of the pain data as acute pain that resolves spontaneously vs. chronic lingering pain cannot be differentiated, where the latter would be more impacted by access and affordability barriers. It is also notable that other aspects of pain experience, (e.g., pain interference) are not available in the dataset, but may have provided additional detail regarding pain impact in combination with pain severity. The AOURP targeted those with lower access to healthcare, in addition to demographic minorities. However, as end users of the data, we cannot confirm to what degree this effort was successful. Given that the AOURP is a volunteer-driven registry, we cannot fully rule out the possibility of healthcare access-based selection bias in the sample. However, other studies of healthcare access and utilization in heart failure using large national datasets are comparable in terms of age (primarily older adults), sex (~ 2% difference in groups across studies), and race (~ 4% difference in groups across studies) [55]. There were discrepancies in missing data by race, which may introduce bias in the race-related results. The larger proportion of missing data for the utilization variables precluded inferential analyses in this study and should be conducted with a more complete sample. The cross-sectional design limits confidence in the observed relationships between the healthcare-related SDOH with 7-day average pain severity, which may be strengthened with analyses of longitudinal datasets [56, 57]. The R² of the final model indicates that variables should be added to account for unexplained variance. For example, some healthcare-related SDOH were not available in the dataset, including health literacy [58], rurality, or Medicaid expansion [59]. Other variables that could constitute important factors include pain phenotypes, medication profiles, HF characteristics (e.g., preserved vs. reduced ejection fraction, stages of HF, New York Heart Association classes) [4], and psychological factors. Diagnosis codes were only obtainable if they were identified by a health provider or present in a linked health record, meaning potential for missed participants in the dataset cannot be fully ruled out.

Conclusion

This study reported healthcare-related SDOH including access, affordability and utilization among adults diagnosed with HF and chronic pain. Participants with higher self-reported access barriers and a higher number of unaffordable services were experiencing significantly higher 7-day average pain severity scores. To our knowledge, this study represents the first examination of comprehensive healthcare-related SDOH among adults with HF and chronic pain, and as such targeted follow-up research is indicated. This research is an important step in understanding the unique barriers to achieving satisfactory pain management in this clinical population.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

AOURP

All of Us Research Program

HF

Heart failure

ICD

International Classification of Diseases

SD

Standard deviation

SDOH

Social determinants of health

US

United States

Author contributions

All named authors meet the criteria for authorship. Dr. Smith created the study design, conducted the analysis, and wrote the manuscript. Drs. Knopf and Emory Moore provided consultation regarding the social determinants of health analysis and revised the manuscript for social determinants of health content. Drs. Zhu and Webster Dekker conducted all statistical and data review from data reports. Dr. Wang provided statistical consultation for the analysis and revised the manuscript for statistical content. Dr. Wierenga oversaw the research study. All authors have read, provided edits for, and approved the final submission.

Funding

No funding was used for this study.

Data availability

This study used data from All of Us Research Programs Registered Tier Dataset v.8. Data are available for authorized users on the Research Workbench. Code used in analyses will be provided upon reasonable request.

Declarations

Ethics approval and consent to participate

All participants provided informed consent before enrolling in the All of Us Research Program. The National Institutes of Health Institutional Review Board approved the All of Us Research Program, and the current study was exempted by the Indiana University Institutional Review Board. The study adhered to the principles outlined in the Declaration of Helsinki.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.