Assessing the Impact of Language Access Regulations on the Provision of Pharmacy Services

Approximately 25 million people in the United States are limited English proficient (LEP). Appropriate language services can improve care for LEP individuals, and health care facilities receiving federal funds are required to provide such services. Recognizing the risk of inadequate comprehension of prescription medication instructions, between 2008 and 2012, New York City and State passed a series of regulations that require chain pharmacies to provide translated prescription labels and other language services to LEP patients. We surveyed pharmacists before (2006) and after (2015) implementation of the regulations to assess their impact in chain pharmacies. Our findings demonstrate a significant improvement in capacity of chains to assist LEP patients. A higher proportion of chain pharmacies surveyed in 2015 reported printing translated labels, access and use of telephone interpreter services, multilingual signage, and documentation of language needs in patient records. These findings illustrate the potential impact of policy changes on institutional practices that impact large and vulnerable portions of the population.


Introduction
Approximately 25 million people in the United States (US) are limited English proficient (LEP) [1]. Compared to English proficient persons, those who are LEP are more likely to have reduced access to care, have a limited comprehension of health information [2], report lower levels of satisfaction with health care services [3,4], and may face higher medical costs due to unnecessary diagnostic testing [3]. Appropriate language services can improve care for LEP individuals [3,4], and health care facilities receiving federal funds, including payments through the Medicaid and Medicare programs, are required-according to Title VI of the Civil Rights Act of 1964 [5,6]-to provide such services [7].
Although language services are increasingly available in many health care settings [8][9][10], several studies have demonstrated inadequate provision of language services in pharmacies. In research conducted in New York City (NYC) in 2006, we found that of pharmacies reporting that they had LEP patients on a daily basis, just 39% reported providing translated prescription medication labels daily and 23% never translated labels. Compared to independent pharmacies, chains were significantly less likely to provide translation services; seven of 65 (10.8%) chain pharmacies in our 2006 study regularly provided translated instructions, compared to 61 of 124 (49.2%) independent pharmacies. More than 80% of all pharmacies included in the study lacked systematic approaches for identification and documentation of linguistic needs and for informing patients of translation capabilities (e.g., language preferences indicated in electronic records, multilingual signage alerting customers to available services) [11]. Studies conducted elsewhere in the US report similar gaps in services [12][13][14].
Recognizing the risks of inadequate comprehension of medication instructions, including poor adherence, errors, and adverse drug events [15], New York lawyers for the Public Interest and Make the Road New York led advocates in New York State (NYS) and NYC in grass-roots campaigns to mandate more systematic and comprehensive approaches to language access services within chain pharmacies. As a result of these campaigns, new regulations were adopted in NYC and NYS between 2008 and 2012 requiring that chain pharmacies provide translation and interpretation services for pharmacy patients speaking the most common non-English languages; however, a system for monitoring compliance by the State Pharmacy Board or other governing body was not included within the regulations. (see Table 5 for a summary of actions and resulting regulations) [16,17].
The goal of this study is to characterize and assess the impact of language access regulations on services available to LEP patients. In this paper, we focus on capacity to provide written translation and oral interpretation of medication instructions, provision of translated materials, and systematic approaches to identifying language needs among NYC chain pharmacies.

Methods
This paper reports on surveys of NYC chain pharmacies, conducted in 2006 and 2015, and focused on translation capacity and practices, as well as observations of pharmacy signage conducted in 2015. It is part of a larger project on language services in chain pharmacies being implemented in NYS, New Jersey, and Connecticut, which includes surveys of chain pharmacies; structured observations of pharmacy signage; and surveys and education of LEP community members.

Data Collection
Study Sample The data used in this analysis come from two random sample telephone surveys of pharmacies: the first conducted in 2006 and the second conducted in 2015. Both surveys used samples selected from current lists of licensed NYS pharmacies maintained by the Office of the Professions, NYS Education Department. Random numbers were generated to select eligible pharmacies. Pharmacies were contacted in sequential order, based on their random number assignment. Trained research assistants conducted all surveys by phone, requesting to speak to a full time, on-duty pharmacist. The research assistants moved to the next pharmacy on the list following a refusal or five unsuccessful attempts to reach a staff pharmacist at the selected pharmacy. Surveys were 5-10 min in length and included responses from pharmacists working during daytime, evening, and weekend shifts.
2015 Survey Because of the interest in the new laws, the 2015 survey covered chain pharmacies only, including 77 in NYC and 77 in other parts of NYS. Eligibility was based on the definitions of a chain pharmacy used in the regulations, which, in NYC, is four or more locations under common ownership. Because we were examining capacity and practice with respect to translation, we limited our sample to pharmacies located in zip codes with relatively high numbers of residents with limited English proficiency. In NYC, this was defined as pharmacies located in zip codes having at least a 10% foreign born population and a population of 10,000 or more, according to the 2014 American Community Survey.
We contacted 154 NYC chain pharmacies to attain our sample of 77 pharmacies (a 50% response rate). We were unable to reach a pharmacist at 52 pharmacies, 23 pharmacies refused to participate (most commonly citing time constraints), and two pharmacies requested a hard copy of the survey instrument but did not complete it. In addition, two pharmacies asked that we seek permission from a district manager prior to survey completion. These potential respondents were excluded, as outreach to district level management was not part of the study protocol.
2006 Survey The 2006 survey included 200 NYC pharmacies, 71 of which were chains, according to self-report. From this sample, we identified 48 chain pharmacies that matched eligibility criteria for the 2015 survey (i.e., located in zip codes with a population of at least 10,000, 10% or more of whom were limited English proficient). The response rate for the 2006 survey was 76% [11].

Study Variables
The 2006 and 2015 surveys included a set of similar questions, the variation reflecting the desire for greater specificity in the latter survey, as well as sensitivity to the language used by pharmacists, for example Bpatients^rather than Bconsumers^(see Appendix 1, Table 6 for relevant questions from the two surveys, according to domain). The surveys focused on (1) pharmacy and pharmacist characteristics and (2) language access capacity and practices that are required by current regulations, specifically: The 2015 survey also asked about knowledge of language access laws in NYS and NYC.
Observations of Signage Recognizing the limitations and potential biases in the self-reported data collected through the telephone surveys, we made structured observations of 30 randomly selected pharmacies in three NYC zip codes where at least 10% of the population were LEP Spanish speakers and the population was at least 10,000. Observations focused on adherence to the signage requirements of the language access laws, namely the presence of signs indicating patients' right to language services, languages included on the signs, location on the pharmacy counter, and sign content.
Community Characteristics Data on LEP prevalence by zip code were obtained from the 2000 Census (2006 pharmacist survey) and the 2014 American Community Survey (2015 pharmacist survey) [18]. These data are available on the US Census website (https://www. census.gov/en.html).

Analysis
Survey Data Frequency distributions were used to describe the pharmacist and pharmacy characteristics and language translation practices. The differences in pharmacy characteristics and practices between 2006 and 2015 were tested for significance with chi-square tests. Using bivariate and multivariate logistic regression, we examined correlates of practices covered by the regulations: (1) translation of medication instructions, (2) access to phone interpretation, (3) language access signage, and (4) indication of language need in patient records. In the multivariate analysis, we adjusted for survey year; pharmacist gender and birthplace; and percent with limited English proficiency in zip code, dichotomized as 10-20% and greater than 20%. Regression analyses were used to estimate odds ratios and 95% confidence intervals. With the exception of the percentage of residents with limited English proficiency in the pharmacy zip code, all data used in the analyses were derived from survey responses. All variables associated at the p < 0.05 in any bivariate analyses were included in the multivariate model.
Observation Data Analysis of observational data focused on proportions with signage at the dispensing counter. Given the small sample size, we did not have significant power for multivariate analyses.
Survey data were analyzed using SAS version 9.3 (Cary, NC). The study protocol was approved by the Institutional Review Board of The New York Academy of Medicine.

Results
Sample Characteristics As shown in Table 1, pharmacists surveyed in 2006 were more likely to be born outside the US and were less likely to report patients with limited English proficiency than were pharmacists surveyed in 2015 (93.8 vs. 100%, p < .05).
Language Capacity and Practice Compared to pharmacies surveyed in 2006, those surveyed in 2015 were significantly more likely to report capacity to print translated prescription medication labels (80.4% in 2006 versus 100% in 2015, p < 0.001), translating labels daily (15.4 vs. 66.7%, p < 0.001), access to a telephone interpretation service (20.8% vs. 90.8%, p < 0.001), and use of the telephone service (20.8 vs. 68.8%, p < 0.001). In addition, there were reported increases in language needs being part of the patient record (19.4 vs. 49.4%, p < 0.01) and written notices or signs indicating access to language services (7.7 vs. 85.7%, p < .001). In 2015, just over 75% of pharmacists reported that efforts had been made to hire bilingual staff, compared to 37.5% in 2006 (p < .001) (see Table 2). Analysis by chain suggests some variability, with greater compliance to the language access regulations at national chains as compared to local chains and pharmacies located within Bbig box^stores and supermarkets (see Table 3). Analysis by language showed the greatest translation capacity for Spanish (98.7%), followed by Chinese (75.3%) and Russian and Italian (68.8% each) (data not shown).
Results from in-person observations conducted in 2015 were relatively consistent with self-report: 80% of observed pharmacies had the required signage at the dispensing counter (data not shown).
In multivariate analyses controlling for pharmacist gender, pharmacist birthplace, and percent LEP in zip code, survey year (2015 compared to 2006) predicted each of the outcomes examined (i.e., requirements of the law): daily translation of medication instructions (AOR 11.7, p < 0.001), access to phone interpretation (AOR 69.7, p < 0.0001), indication of language need in patient records (AOR 6.8, p < 0.001), and language access signage (AOR 131.1, p < 0.0001(see Table 4).

Discussion
Our findings demonstrate that between 2006 and 2015 there was a significant improvement in the capacity of NYC chain pharmacies to provide language access services, consistent with NYC and NYS regulations adopted between 2008 and 2012 (Table 5). A significantly higher proportion of chain pharmacies surveyed reported being able to print translated medication labels, providing translated labels to LEP patients, access and use of telephone interpreter services, multilingual signage informing patients of language services, and documentation of language needs in patient records-all of which are requirements of the new regulations and laws   adopted in New York. These findings illustrate the potential impact of grass-roots advocacy and resultant policy change on institutional practices that impact significant portions of the population. Our ability to conduct surveys using similar protocols before and after the implementation of pertinent regulations and laws is a particular strength of this study. The significant change in chain pharmacy practices related to language access over just a few years is notable, given early concerns related to feasibility and liability exposure, which were expressed by pharmacists when we conducted our earlier study [11].
As far as we know, New York's pharmacy language regulations are unique. The findings reported here could thus be useful in the development of similar policy efforts in other jurisdictions. Most of the pharmacies included in this study are part of multistate chains that have developed systems for the provision of language services. Promoting and utilizing those systems outside of New York, seemingly, could be accomplished with relative ease.
This study has several limitations. At both points in time, data were largely self-reported, and were therefore subject to recall errors and social-desirability bias. Although the reliability of recall should be similar for the 2006 and 2015 surveys, it is possible that, given the change in regulations, social-desirability bias may have been stronger in 2015. In addition, the pharmacist perspective represents just a part of the story: receipt of language-appropriate instructions by patients with limited English proficiency, improved comprehension, reduced medication errors, and better health are the ultimate goals. We have been surveying Spanish-speaking LEP patients as part of the larger study, to better understand their experience with language services in chain pharmacies and the implications thereof. Finally, similar research on chain pharmacies in other states is necessary to verify that the changes reported here were the result of New York regulations, rather than more general shifts in pharmacy practice. We are carrying out surveys and structured observations in communities in Connecticut and New Jersey (contiguous states to NYS) with high numbers of LEP residents to address this concern.
Despite these limitations, the significant changes documented in this study were largely consistent across the specific language services covered by the regulations. The results make a convincing case for the effectiveness of the laws and regulations implemented and for their adaptation in other states and municipalities.