Do state-level cultural competency educational mandates influence patient experience?

In our diverse society, healthcare providers have recognized that patient-centered care and cultural competency are essential for quality healthcare. However, navigating this landscape requires specialized knowledge, forward-thinking, meaningful community engagement, and an understanding of America’s changing demographics. Hospitals and health systems face complex challenges, including connecting with vulnerable populations, building and maintaining a culturally competent workforce, and adapting to evolving regulations. Delivering high-quality care goes beyond general cultural awareness; it requires cultural sensitivity based on a specific understanding of different communities’ values, beliefs, and practices. In other words, developing a high level of cultural competence within healthcare organizations is essential.

The importance of cultural competency in the United States has come a long way from its origins in the 1960s and 1970s, when the civil rights movement focused on racial and ethnic inequalities with an ultimatum for respect of all cultures. The seminal text Crossing the Quality Chasm defined patient-centered care as ‘providing care that is respectful of, and responsive to, individual patient preferences, needs, and values, and ensuring that patient values guide all clinical decisions.’^[1] This was followed shortly thereafter by the Office of Minority Health in the Department of Health and Human Services (DHHS) laying the foundation to address diverse populations with the development of the Culturally and Linguistically Appropriate Services (CLAS) standards in December 2000.^[2] The standards encompass governance, leadership, and workforce; communication and language assistance; and engagement, continuous improvement, and accountability. These standards are rooted in culturally competent care that is seen as a basis for reducing disparities.^{[2, 3]} In 2009, the National Quality Forum issued a consensus report presenting a comprehensive framework and preferred practices for measuring and reporting cultural competency across healthcare organizations.^[4] Cultural differences represent a significant and ongoing challenge for healthcare providers striving to deliver high-quality, equitable care. These disparities are often viewed as race or ethnicity; however, the range is much broader to include gender, race, social, cultural, gender identity and orientation, religion, disability, and linguistic barriers.^[5]

These policy efforts underscore a growing recognition that health equity cannot be achieved without intentional strategies to acknowledge and address cultural differences through education, workforce development, and systemic change. Culturally responsive education and training are essential as the U.S. healthcare system becomes more inclusive and patient-centered. Equipping healthcare professionals with the knowledge, skills, and attitudes necessary to navigate cultural diversity is critical to reducing disparities, improving communication, building trust, and enhancing health outcomes for marginalized and vulnerable populations.

Recently, laws and new mandates related to cultural competency training for healthcare professionals have been enacted. For example, effective January 1, 2025, cultural competency training of healthcare providers in Illinois commenced. Other states, such as California, Connecticut, and New Jersey, already had mandatory cultural competency training. Cultural competence is widely promoted as a strategy to reduce health inequities, and training has accordingly become a central policy lever in healthcare organizations.^[6] A systematic review of thirty-four provider education interventions found consistent beneficial effects on patient satisfaction, though evidence for downstream health outcomes was limited.^[7]

This research examines healthcare organizations in states that have enacted cultural competency education legislation. We compare patient satisfaction data (i.e., Hospital Consumer Assessment of Healthcare Providers and Systems [HCAHPS] scores) before and after implementing the education mandates. In addition, to provide a broader and more balanced perspective, the research includes a comparative analysis against two groups of states without cultural competency education legislation: five matched comparison states and the remaining states. This dual approach assesses the association between cultural competency training mandates and patient satisfaction.

1.1Cultural competency education legislation

1.2Connecticut’s implementation of cultural competency education mandate

1.3New Jersey’s implementation of cultural competency education mandate

1.4Oregon’s implementation of cultural competency education mandate

1.5Washington’s implementation of cultural competency education mandate

To answer our research question, “Do patients receiving care in hospitals in states with a cultural competency educational mandate report greater improvement in HCAHPS scores between 2018 and 2023 compared to states with no mandate for cultural competency training?” We employed the 2018 and 2023 HCAHPS surveys. The final sample consisted of state-level HCAHPS scores from 48 states (Nevada and New Mexico were excluded; see narrative below) in the United States and the District of Columbia for 2018 and 2023. To do this, we examined changes in the state-level star ratings between 2018 and 2023 for nine HCAHPS measures of patient experience using t-tests.

HCAHPS scores are calculated based on the quarterly patient survey responses and are reported at the hospital, state-wide, and national level.^[25] We used state-level HCAHPS for the current analysis. Each hospital’s score in each state on a particular measure is averaged, and a state HCAHPS score for that measure is then calculated.^[26] The study examined nine HCAHPS measures of patient experience. These measures were chosen as they are more direct representatives of culturally competent care. The other components of HCAHPS include ratings on factors such as cleanliness and noise, which are less impacted by cultural competence. This selection aligns with the provider-focused nature of cultural competency mandates across all five states, which primarily targeted individual healthcare professionals’ training rather than institutional or environmental factors. The reliability and validity of each care composite measure have been extensively evaluated in prior research.^{[26, 27]}

As noted above, consistent with the original study design, Nevada and New Mexico (which adopted mandates in 2005 and 2007) were excluded from all groups. We divided the remaining 48 states and the District of Columbia (DC) into three groups: five focal states that have enacted cultural competency training mandates (California, Connecticut, New Jersey, Oregon, and Washington; enacted and effective dates in Table 2), hereafter the mandate states, five states that were matched to the five cultural competency training mandated states based on similarity index scores (New York, Delaware, Illinois, Minnesota, Michigan), and the remaining 38 states and the District of Columbia. The comparison states were selected at the study’s inception as a group, rather than through one-to-one pairing, with reference to the State Similarity Index, a publicly available composite that equally weights demography, culture, politics, infrastructure, and geography.^[28] The five were judged collectively similar to the mandate states as a set, and pair-level index values were therefore not part of the design. The adequacy of the grouping is further evaluated empirically, through the baseline equivalence tests reported below and through an alternative matching. Each mandate state was matched to one comparison state, without replacement, by the assignment minimizing total Euclidean distance on six standardized covariates measured at or nearest the 2018 baseline: hospital beds per 1,000 population (2018), the for-profit share of community hospitals (2018), urban population share (2020 census), Medicaid expansion status in effect on January 1, 2018, per-capita personal health care spending (2020), and the 2018 HCAHPS overall star rating. The candidate pool comprised the 42 eligible non-mandate states, excluding Nevada and New Mexico per the original design, the District of Columbia, and Illinois given its 2025 mandate. The resulting pairs are California-Rhode Island, Connecticut-Delaware, New Jersey-Maryland, Oregon-Hawaii, and Washington-Michigan. All analyses were repeated with this alternative matched group as a sensitivity analysis. We employed paired t-tests to assess whether each group’s state-level star ratings changed significantly between 2018 and 2023. To compare groups directly, we also estimated a one-way ANOVA on the percent change in each measure across the three groups, with Bonferroni-corrected pairwise contrasts and Holm correction across the nine measures. Kruskal-Wallis tests were estimated as distribution-free robustness checks. Because HCAHPS star ratings are normed to national performance, we repeated the between-group comparison using top-box and linear-mean scores. Group-level results are presented in Table 3, and between-group results in Table 4. Sensitivity analyses excluding Washington and excluding Illinois, and analyses across alternative year-windows, assessed robustness.

Table 3 presents the results of the paired t-tests evaluating changes in HCAHPS between 2018 and 2023 for each group. For all HCAHPS measures, increases in star ratings represent improvements in patient experience while decreases represent declines in performance. Among the five mandate states, there were significant increases in their star ratings for doctor communication (p < .001), communication about medicines (p < .05), and discharge information (p < .05). The doctor communication improvement in mandated states paralleled improvements in the remaining states. However, mandated states experienced significant declines in nurse communication scores (p < .05) and overall hospital ratings (p < .05). No significant changes were observed in staff responsiveness, care transition, willingness to recommend, or summary star ratings, with the latter maintaining complete stability (3.01 to 3.01).

A one-way ANOVA of the percent change from 2018 to 2023 across the three groups (see Table 4) showed no measure on which mandate states improved more than comparison states. Overall hospital rating declined more in the mandate and matched groups than in the remaining group [F(2,46) = 7.18, p = .002, surviving Holm correction]; nurse communication showed the same direction (F = 4.57, p = .015) but did not survive family-wise correction. Mandate and matched groups did not differ on any measure. Because star ratings are normed nationally, the comparison was repeated on top-box and linear-mean scores (see Table 5), which confirmed the pattern on overall hospital rating (top-box F = 6.72, p = .003; linear-mean F = 10.74, p < .001). The overall-rating result was robust to excluding Washington and to excluding Illinois on all three metrics, while the nurse-communication star contrast attenuated without Washington (see Table 6); on the absolute metrics, the overall-rating difference was significant in all four year-windows examined (see Table 7). Baseline (2018) levels did not differ significantly across groups (see Table 8). Under the covariate-based alternative match, whose pairs, distances, and covariates are reported in Table 9, the between-group results were unchanged in structure and modestly stronger: the overall-rating contrast persisted [F(2,46) = 6.56, p = .003, surviving Holm correction], the nurse-communication contrast now also survived family-wise correction (F = 6.72, p = .003, Holm p = .025), and mandate and alternative-matched states differed on no measure (see Table 10).

The five matched states demonstrated significant increases only in discharge information (p < .05) and significant declines in overall hospital ratings (p < .05). Unlike the states with cultural competency training mandates, the matched states showed no significant changes in nurse communication, doctor communication, staff responsiveness, communication about medicines, care transition, willingness to recommend, or summary star ratings. The remaining 38 states and the District of Columbia showed significant increases in doctor communication (p < .001), communication about medicines (p < .001), discharge information (p < .001), willingness to recommend star ratings (p < .001), and summary star ratings (p < .05). Similar to mandate states, the remaining states and the District of Columbia had a significant decline in nurse communication (p < .05). Additionally, this group experienced significant declines in staff responsiveness (p < .01) and overall hospital ratings (p < .01). No significant changes were observed in care transition (3.18 to 3.18) in this group.

Notably, willingness to recommend showed divergent patterns across groups: significant improvement only in the remaining states and the District of Columbia (p < .001), while both mandated states and matched states showed no significant change. Similarly, staff responsiveness demonstrated a significant decline only in the remaining states and the District of Columbia (p < .01), while mandated and matched states showed non-significant declines. Care transition demonstrated no significant changes across any of the three groups.

All three groups experienced declines in overall hospital ratings, with mandated states declining from 3.17 to 2.91 (p < .05), matched states from 3.21 to 2.92 (p < .05), and the remaining states and the District of Columbia from 3.37 to 3.29 (p < .01). Summary star ratings improved significantly only in the remaining states and the District of Columbia (from 3.34 to 3.39, p < .05), remained stable in mandated states (3.01 to 3.01), and showed no significant change in matched states (3.13 to 3.15).

The direct between-group tests reframe the study’s central finding. No HCAHPS measure showed mandate states improving more than either comparison group; on overall hospital rating, mandate and matched states declined more than the remaining states, a pattern that held on the absolute top-box and linear-mean metrics, while mandate and matched states did not differ from each other on any measure (see Tables 4 and 5). The differential patterns of improvement across domains and between groups reveal both the potential benefits and limitations of cultural competency mandates.

The significant improvements in doctor communication across mandated states (p < .001) and the remaining states and the District of Columbia (p < .001), contrasted with non-significant changes in matched states, are consistent with cultural competency training supporting enhanced physician-patient interactions. The parallel improvements between mandated states and the remaining states indicate that mandated states kept pace with broader national trends in this domain.^{[29, 30]} On the absolute metrics, however, doctor-communication scores declined in all three groups (see Table 5), indicating that the star-rating improvements partly reflect national re-norming of the star scale rather than rising absolute performance.

Mandated states improved significantly in communication about medicines (p < .05), while their matched comparison states did not. This gain was not unique to the mandate group, however; the remaining states improved in this domain as well (p < .001). Many cultural competency programs emphasize linguistic competency and health literacy,^[10] but the broad improvement across non-mandated states indicates that this gain cannot be attributed to the mandate alone.

The universal improvements in discharge information across all three groups, mandated states (p < .05), matched states (p < .05), and the remaining states and the District of Columbia (p < .001), suggest broader systemic initiatives targeting care transitions and patient education that extended beyond state-level cultural competency policies. However, the magnitude and significance levels varied, with the remaining states showing the strongest improvements (p < .001), potentially reflecting synergies between multiple quality improvement initiatives.

The divergent patterns in willingness to recommend merit particular attention. The significant improvement observed only in the remaining states and the District of Columbia (p < .001), with no change in mandated or matched states, suggests that cultural competency mandates alone were not associated with willingness to recommend, which may instead relate to overall service quality, facility amenities, and market competition.^[31]

The universal decline in nurse communication scores, significant in mandated states (p < .05) and the remaining states and the District of Columbia (p < .05), with non-significant decline in matched states, presents a complex interpretive challenge. This pattern suggests broader systemic challenges in nursing practice during the study period, including workforce shortages, pandemic-related stress, and evolving care delivery models, which are not specific to state-level policy interventions. The non-significant decline in matched states is consistent with regional variation in nursing workforce challenges.^[32] The decline in nurse communication is also difficult to attribute to the mandates themselves, since the mandates reach registered nurses unevenly: California addresses nurses only for implicit-bias content, Connecticut only advanced practice nurses, and New Jersey only perinatal nurses, whereas Oregon and Washington include registered nurses directly and fully (see Table 2). A patient-experience domain driven by bedside nursing would not be expected to respond uniformly to mandates with such heterogeneous nurse coverage.

The decline in staff responsiveness observed only as significant in the remaining states and the District of Columbia (p < .01), while mandated and matched states showed non-significant declines, presents an unexpected pattern. This finding suggests that broader systemic factors affecting staff responsiveness may have differentially impacted states based on their size and diversity, with the larger group of remaining states experiencing more pronounced challenges.

The absence of significant changes in care transition across all groups reveals domains particularly resistant to improvement through current intervention strategies. These findings suggest that neither cultural competency mandates nor other concurrent quality initiatives successfully addressed these aspects of patient experience during the study period. The universal lack of improvement in these domains is consistent with structural healthcare delivery challenges that require more comprehensive system-level interventions.^[33]

The stability of summary star ratings in mandated states, maintaining identical scores (3.01 to 3.01) during a period that included the COVID-19 pandemic, contrasts with the significant improvement in the remaining states (p < .05). This pattern, alongside the non-significant change in matched states, is consistent with mandates providing stability rather than driving improvement, potentially reflecting a floor effect or the challenge of achieving gains in states already focused on cultural competency.^[34]

The decline in overall hospital ratings observed in all three groups, significant in mandated states (p < .05), matched states (p < .05), and the remaining states and the District of Columbia (p < .01), suggests that broader systemic factors affected overall hospital experiences during the study period. The universal nature of these declines suggests that factors beyond cultural competency training shaped overall hospital experiences.^[35]

Counting the number of domains in which each group improved is not a valid basis for comparison, because the remaining group contains far more states and is therefore better powered to reach significance at the same effect size. The appropriate test is the direct between-group comparison reported above, which found no measure on which mandate states improved more than either comparison group. The domain-count difference should therefore not be read as evidence that broader quality initiatives outperform cultural competency mandates.

The implementation variations across mandate states (see Table 1) highlight important differences in how cultural competency training requirements were operationalized. Given the heterogeneity of these mandates in professions covered, hours, content, and timing (see Table 2), and the null mandate-versus-matched contrasts on every measure, the present design cannot attribute outcomes to either the presence or the structure of a mandate. Prior empirical work has linked cultural competence training to improved patient satisfaction at the level of individual provider education;^{[7, 36]} whether that provider-level benefit scales to state-mandated training measured through population-level HCAHPS remains an open question that these findings do not resolve.

Descriptively, the within-group changes in the mandate states were concentrated in provider communication domains, with no change in system-level metrics such as care transition and staff responsiveness, a pattern consistent with the individual-focused nature of most cultural competency training programs, though the between-group tests do not attribute it to the mandates. Without addressing organizational structures and system-level processes alongside individual training, the full potential of cultural competency mandates may not be realized in broader patient experience domains.

4.1Limitations

A direct between-group test found no measure on which mandate states improved more than comparison states. However, universal declines in nurse communication and lack of improvement in staff responsiveness and care transition indicate broader systemic challenges unaffected by cultural competency policies. These exploratory findings indicate that mandates were not associated with better patient experience: mandate and matched states declined similarly, and both declined more than the remaining states. Because HCAHPS is a single patient-experience outcome and the mandate groups are small and heterogeneous, the question merits further study with hospital-level designs and broader outcomes before policy recommendations are made.

MA: data analysis and interpretation, manuscript preparation, manuscript editing, manuscript review. DU: manuscript editing, manuscript review. RD: conception, manuscript editing, manuscript review. PC: manuscript editing, manuscript review. KRH: statistical analysis, data analysis and interpretation, manuscript review. NB: design, data acquisition, supervision, manuscript review. All authors have given final approval for the version to be published.

None.

The authors declare they have no conflicts of interest.

Not applicable. The study used publicly available, aggregate data and did not involve human subjects.

The ethical statement is not applicable. This study used publicly available, aggregate data and did not involve human subjects.

The Publication Ethics Committee of the Association for Health Sciences and Education. The journal’s policies adhere to the Core Practices established by the Committee on Publication Ethics (COPE).

Not commissioned; externally double-blind peer reviewed.

The datasets analyzed during the current study are publicly available: state-level HCAHPS data from the Centers for Medicare & Medicaid Services (CMS); hospital beds, hospital ownership, and health spending data from KFF State Health Facts and the CMS State Health Expenditure Accounts; and urban population data from the U.S. Census Bureau (2020 Census).

Derived datasets and analysis code are available from the corresponding author on reasonable request.

The authors used (Claude AI) as a language-editing tool to improve grammar, clarity, and readability. The AI tool was not used for data analysis, data generation, image processing, or scientific content creation. All outputs were carefully reviewed and verified by the authors, who take full responsibility for the final content.