The main finding of this study was that increased social deprivation was associated with decreased outpatient satisfaction, as measured by the Press Ganey Outpatient Medical Practice Survey. This observation was observed for both the Press Ganey Total Score and the Provider Sub-Score, and was independent of several factors previously shown to have a large magnitude of impact on patient satisfaction including wait time, patient age, sex and race (41, 55, 57, 58). The secondary findings of the study are in-line with much prior work on patient satisfaction and associated factors such as patient age, sex, race, insurance status, wait time, provider type, and specialty type.(41, 55, 57, 58)
The importance of understanding how socioeconomic factors affect the utilization of and access to the healthcare system is becoming increasingly evident. The role of a patient’s social and economic circumstances in their overall physical and mental health has been well-elucidated (3, 11, 31, 46, 48, 53, 58-60), and our findings are consistent with a limited number of previous studies in documenting an association between socioeconomic status and patient satisfaction scores.
Young et al. demonstrated that average income levels based on zip codes and lower patient satisfaction scores are correlated among elderly patients seen in various specialty clinics (61). McFarland et al. evaluated 934,000 patients and showed Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) survey scores were directly correlated with education level, which has been used as a surrogate for socioeconomic status in the literature (38). Additionally, Nieman et al. demonstrated that lower socioeconomic status correlated with worse PGOMPS scores among the pediatric surgical population (40).
Critically, this study does not address whether the quality of health care differed based on a patient’s social deprivation. The PGOMPS does not measure health care quality, but rather satisfaction with the process of outpatient health care provision. It is important to make the distinction between quality and satisfaction, particularly when interpreting evidence suggesting that social deprivation may influence a patient’s perception of the care they received (22, 62-65). Along those lines, this study does not address the underlying causes of differences in patient satisfaction scores between patients with different levels of social deprivation. Arpey et al. demonstrated that patients of lower SES were more likely to perceive their economic status as influencing their care than those of higher socioeconomic brackets (62). Verlinde et al. conducted a systematic review evaluating how SES affects physician patient communication. In their study they found that patients of lower SES are less likely to participate in shared decision making and ask questions (66). Schroder et al. pointed out that patients in the bottom socioeconomic quartile had less medical knowledge and were less likely to desire to play an active role in their disease management than patients in the highest quartile (67). Studies by Wright et al. and Okoroafor et al. demonstrated that patients were more likely to report high levels of anxiety and depression in worse social deprivation indexes (30, 31), and Tyser et al. and Tisano et al. independently showed that patients with worse PROMIS anxiety and depression scores were less likely to report satisfaction on the PGOMPS (58, 68). An additional study by Schroder et al. found that patients of lower SES were more likely to wait to seek care for their heart disease until after they suffered a myocardial infarction (69). Previous literature has also shown that physicians who treat patients with greater disease severity and worse prognosis are more likely to receive lower patient satisfaction scores (70-73).
Objective discrepancies of care based on socioeconomic status have, however, been documented. Govindarajan and Schull found that patients residing in economically deprived neighborhoods were less likely to have advanced paramedic teams dispatched to their location and had greater transport time to hospitals when controlled for distance than those residing in less economically deprived neighborhoods (24). Patel et al. found that time between initial encounter for an ACL tear to surgery was greater for pediatric patients from lower socio-economic settings (23). The systematic review conducted by Verlinde et al. also found that lower SES patients received less overall communication and fewer explanations directed to their understanding level than those of higher SES (66). It remains uncertain if these inequalities of care are a result of limited access of care and insurance/payment difficulties, or rather due to inherent biases. Clearly, the interplay between socioeconomic status, social deprivation, and the healthcare delivery process is complex. Further work is needed to evaluate for and potentially reduce discrepancies of care that these patients may experience in line with the overall goal of providing equitable and high-quality care.
The Patient Protection and Affordable Care Act enables Medicare to make incentive payments to hospitals based on specific quality domains that include the patient experience of care, and have in turn been used to adjust physician compensation (74, 75). Our findings may also help inform health care policy makers and/or administrators in decision making surrounding attaching patient satisfaction scores to various methods of reimbursements. The impact of such policies should be evident: without accounting for the impact of a patient’s economic disadvantage on satisfaction scores, providers who have reimbursement tied to satisfaction scores may be disincentivized from caring for patients with, or working in areas with, high levels social deprivation. This could further perpetuate the disparities that these policies are attempting to correct. An example of this was demonstrated in a study that evaluated the impact of the 2019 peer group stratification of Medicare’s Hospital Readmission Reduction Program (HRRP) in the United States. The HRRP allows for a penalization to be enacted if hospitals have readmission rates greater than 30 days (76). In 2016, The United States Congress passed the 21st Century Cures Act allowing HRRP to take into consideration the effect of social deprivation on readmission rates (77). Under HRRP in 2019, hospital performance was stratified into quintiles based on patient socioeconomic status and the proportion enrolled in Medicare and Medicaid. The cost of readmission penalties to hospitals and subsequently physician reimbursements were cut in half for hospital’s in the most deprived quintile as demonstrated by Joynt Maddox et al (78). The importance of accounting for SES in evaluation of health care quality has also been demonstrated outside of the United States.(79)
There are several limitations of this study. The generalization of our findings to other health care systems with differing regional and patient demographics is limited given that our study was conducted at a single institution treating a population that is predominately white. Furthermore, our institution provides care for patients from a large geographical distribution. Many patients, often from underserved and economically disadvantaged areas, travel up to several hours to be seen by specialists at our institution. The expectations, and therefore satisfaction, of these patients be different from other hospital systems with smaller catchment areas. Our study is also limited by a non-response bias, which is also an inherent limitation of the PGOMPS in general. Previous literature from our institution has shown the PGOMPS response rate to ranges from 8.9 to 16.5% (39, 41). Tyser et al. found that responders differed from non-responders in terms of age, sex, and insurance type (56). These factors are a real-world limitation of PGOMPS, and should not only be taken into account when interpreting study results, but also when determining the applicability of the survey as a determinant of vale of care and reimbursement rates. Although we only included new patient visit patient encounters, it is possible that a patient’s economic situation, and satisfaction with care, could potentially change throughout a treatment course. Lastly, the magnitude the association between ADI and satisfaction are seemingly small in comparison to patient age and wait time, but the effects are additive for increasing deciles of social deprivation and the comparison between highest and lowest quartiles demonstrates a significant difference.