Assessing Spanish Health Literacy and Cervical Cancer Knowledge, Attitudes, and Behaviors in a Student-Run Free Clinic

While the incidence of cervical cancer continues to decrease, there is a significant discrepancy in incidence rates and screening behaviors among Hispanic and non-Hispanic white patients in the USA. This project examines the relationship between Spanish health literacy and cervical cancer screening knowledge, attitudes, and practices among native Spanish-speaking patients at risk for cervical cancer at the USF BRIDGE Healthcare Clinic, a student-run free clinic in Tampa, FL. Spanish-speaking patients ≥21 years (n = 34) participated in a quality improvement project that included an assessment of Spanish health literacy and a written survey on cervical cancer knowledge. Chi-squared tests were performed to assess potential relationships between health literacy and cervical cancer knowledge, attitudes, health behaviors, and demographics. Seven participants (20.6%) scored between 0 and 14 on the SAHL-S, indicating inadequate health literacy. A significant difference in cervical cancer health knowledge was found between patients with adequate health literacy compared to patients with inadequate health literacy (p = 0.002). There is a potential association between low Spanish health literacy and subsequent poorer understanding of cervical cancer in BRIDGE patients. This implies that patients of low health literacy may have poorer comprehension of other aspects of their care beyond cervical cancer screening. Strategies are discussed to improve communication with BRIDGE patients of low Spanish health literacy that may be applicable to other patient populations.


Introduction
The American Cancer Society (ACS) estimates that about 14,100 new cases of cervical cancer will be diagnosed in the USA in 2022 [1]. While cervical cancer is the fourth most frequently diagnosed cancer in women worldwide, incidence rates in the USA have decreased by more than half since the 1970s and continue to decline by approximately 0.8% per year [1]. The aforementioned trends are largely due to universal adoption of preventative practices including Papanicolaou (Pap) testing and human papilloma virus (HPV) vaccination [1]. However, there is a significant discrepancy in incidence rates between different races and ethnicities in the USA [2,3].
According to a 2009 analysis, the relative risk of being diagnosed with invasive cervical cancer in the USA was 1.8 times higher for Hispanic women than non-Hispanic white women from 2000 to 2004 [2]. In 2018, the incidence of cervical cancer was higher for Hispanic women than for any other racial or ethnic group with approximately 9.6 new cases per 100,000 women [4]. Variations in cancer screening utilization and socioeconomic status (SES) are thought to account for much of this observed difference. According to the American Cancer Society, 83% of Hispanic women ages 25-65 years were up to date with screening in 2018 compared with 86% of non-Hispanic white women [5]. However, multiple studies have demonstrated that Hispanic women of lower SES are less likely to participate in routine screening [3]. There is also evidence to suggest that followup of abnormal results may be different in Hispanic women, which further indicates other factors at play such as acculturation, lack of education, language barriers, and SES [6]. Additionally, a 2007 study indicated that 32.0%, 20.7%, and 20.2% of recently emigrated women from Mexico, Central America, and South America respectively had never had a Pap test as opposed to 5.4% of US-born women [7]. Puerto Rican women seem to have the lowest rates of cervical cancer screening at a prevalence of 80% [5], which is thought to be related to Puerto Rico's multiple recent environmental disasters and public health emergencies [8].
Studies have shown that screening avoidance in Hispanic women is attributed to fear, embarrassment, and fatalistic attitudes toward cancer [9]. Another commonly observed barrier is the perception that screening for cervical cancer is unnecessary if one is asymptomatic and without a family history of disease [9].
In a study of Spanish-speaking, low-income Latinas aged 40-78, participants with low health literacy were significantly less likely to have ever had a Pap test than individuals with adequate health literacy even after controlling for factors such as health insurance, education, and SES [10]. This indicates that even when health information is provided in Spanish, inadequate health literacy, defined by Healthy People 2030 as an individual's ability to use information/ services to make health-related decisions/actions, contributes to the widening disparity in cervical cancer knowledge and screening behaviors [11]. A prospective cohort study of English-speaking patients similarly notes the association between inadequate health literacy and cervical cancer screening knowledge [12].
Examination of health literacy has the potential to illuminate further aspects of the disparity in cervical cancer incidence and screening behaviors in the Hispanic population. Our project examines potential relationships between Spanish health literacy and cervical cancer screening knowledge, attitudes, and practices in the patient population of the USF BRIDGE Healthcare Clinic, a student-run free clinic that primarily serves Spanish-speaking patients in the Tampa area. Following this analysis, we hope to implement targeted interventions to address specific gaps in care that we may find regarding health literacy or cervical cancer screening in our BRIDGE patients.

Methods
From November 2021 to March 2022, eligible patients were identified and recruited on approximately 2-3 in-person clinics monthly. Inclusion criteria included being a BRIDGE patient, having or previously having a cervix, speaking Spanish as a native language, and being over 21. All BRIDGE patients are uninsured with a household income below the national poverty line. A preliminary search of BRIDGE records indicated 138 eligible patients as of October 2021. Since this was a quality improvement project classified by USF IRB as Not Human Subjects Research, the project did not require IRB approval.
Our investigation utilized a cross-sectional design. Participation included administration of a previously validated health literacy test and a written questionnaire to assess patients' cervical cancer knowledge and attitudes. All materials were written in Spanish and administered by native Spanish speakers. Trained student investigators approached patients in their exam rooms after they were seen by providers.
After obtaining verbal consent, participants' Spanish health literacy was measured by the SAHL-S, a verbal assessment designed to assess Spanish-speaking adults' reading and comprehension of common medical terms [13]. To ease administration of the assessment, our project utilized 4″ by 5″ flashcards as recommended in the SAHL-S User Guide. For each stem, participants were shown the flashcard and asked to read the stem and associated words. They were then instructed to pick which of the key or distractor words was most similar to the stem. A correct answer for each stem included both correct pronunciation and accurate association. Each correct response counted as one point. A score between 0 and 14 suggested inadequate health literacy. A score above 14 indicated adequate health literacy.
After completing the SAHL-S, participants' cervical cancer knowledge and attitudes were assessed by a written questionnaire. Survey items were based on those from previous studies on cancer health literacy and included demographics, knowledge statements, attitude statements, and open-ended questions (Lindau et al. 2002). Demographic questions elicited patients' highest level of education, level of English understanding, and birthplace. Six knowledge statements surveyed patients' understanding of the association between HPV, cervical cancer, and Pap tests. Three attitude statements elicited patients' views regarding regular Pap tests and abnormal Pap test results. Seven of the nine total statements were assessed using the 5-point Likert scale. Two of the nine statements required a yes or no response. One open-ended question asked participants to describe any barriers they faced when seeking cervical cancer screening while the second elicited any additional comments or feedback. The questionnaire was developed in English by all of the project investigators and then translated into Spanish by one individual. The questionnaire was then translated back into English by a separate individual and compared to the original English survey, where only one insignificant grammar difference was found.
A Qualtrics survey was developed to record and store patients' SAHLS scores and questionnaire responses. Each participant's data was stored using the patient's initials and the date they were recruited. Once the patient recruitment period was over, these identifiers allowed project investigators to obtain additional demographic information and cervical cancer screening behaviors from the electronic medical record. US Preventive Services Task Force recommendations were utilized to determine if patients' cervical cancer screening behaviors were adequate or inadequate [14].
Data was translated into English and transferred from an excel sheet to IBM SPSS. Transfer of data was reviewed independently by two authors to ensure quality control. Participant responses were quantified using descriptive statistics. This included frequencies, percentages, means, and standard deviations. Demographics, including SAHL-S scores, were listed in tabular form. Participants' responses to the cervical cancer knowledge and attitudes questions were presented graphically.
To objectively assess participants' cervical cancer knowledge, the six knowledge questions were each assigned a numerical value. For the two yes/no questions, a score of 1 was assigned when the correct answer was selected and a score of 0 was assigned otherwise. For the four Likert scale questions, a score of 1 was assigned if they correctly agreed with the answer, either selecting "strongly agree/disagree" or "agree/disagree." A score of 0 was assigned for neutral/ incorrect responses. The sum of these six numerical values created our cervical cancer knowledge variable, with possible scores of 0-6. Those who scored greater than 4 points were assigned the categorical label of adequate cervical cancer knowledge. Those who scored 4 points or less were assigned the label of inadequate cervical cancer knowledge.
A chi-squared test was conducted to determine whether there was a difference in cervical cancer knowledge between those with adequate vs inadequate health literacy. Further chi-squared tests were run to examine potential relationships between health literacy and specific aspects of cervical cancer knowledge, attitude responses, and demographics.
Qualitative analysis was performed for the two freeresponse questions using thematic deductive content analysis [15]. Responses to the two free-response questions were coded in relation to the question prompt, summarized by one coder, and reviewed by a second coder. Discrepancies were resolved through discussion. A word cloud was generated using the Google Docs Word Cloud add-on to graphically display relevant representative quotes.

Demographics
Of the 38 patients who were approached, 4 declined to participate. Thirty-four patients completed the SAHL-S as well as the cervical cancer questionnaire ( Table 1). The average age of participants was 46.3 with a standard deviation of 11.4. The majority of participants were married or partnered (79.4%), have never smoked (91.2%), born in Honduras or Mexico (64.9%), had a high school education or higher (58.9%), and understood only basic phrases in English (55.9%). In terms of cervical cancer screening practices, the majority of participants had received a Pap test in the past (97.1%) and were up to date on their Pap test screening (82.4%). Twenty-seven participants (79.4) scored above a 14 on the SAHL-S, indicating adequate health literacy. Seven participants (20.6%) scored between 0 and 14, indicating inadequate health literacy.

3
Chi-squared tests of independence were run to investigate potential associations between the above demographics. It was determined that patients with a middle school education or higher were more likely to have adequate health literacy (p = 0.019). No significant associations were found between health literacy and marital status, tobacco use, birthplace, high school education or higher, or cervical cancer screening practices (Tables 2 and 3).

Cervical Cancer Knowledge
Most patients had heard of cervical cancer concepts (Fig. 1); 26 participants (76.5%) had heard of cervical cancer and 32 (94.1%) had heard of Pap tests. Twenty six participants (76.5%) indicated they strongly agreed or agreed with the statement that HPV causes cervical cancer. Thirty-two respondents (94.1%) correctly agreed that a Pap test samples cells from the cervix to check for risk of cervical cancer. "Neutral" was the most common response to the statement that Hispanic women have higher rates of cervical cancer than non-Hispanic women, indicating lower knowledge of this disparity. Nearly all participants (33 of 34) agreed that getting regular Pap tests can prevent cervical cancer.
Individual answers to the knowledge questions were examined via chi-squared tests to see if there were any associations with the demographics described above. It was determined that patients with inadequate health literacy, meaning a SAHL-S score between 0 and 14, were more likely to have never heard of a Pap test (p = 0.004). Additionally, patients with a middle school education or higher were more likely to have heard of a Pap test (0.009).
Participants' overall cervical cancer knowledge scores were determined via combining the six knowledge questions as described in the methods and ranged from 2 to 6. The average knowledge score of participants with adequate health literacy was 5.07, with SD of 1.2. The average score of participants with inadequate health literacy was 4.14, with SD of Nine participants (26.5) missed more than 2 of the knowledge questions, receiving a label of inadequate cervical cancer knowledge. A chi-squared test illustrated a significant difference between cervical cancer health knowledge in patients with adequate health literacy compared to patients with inadequate health literacy (p = 0.002). No significant associations were found between cervical cancer knowledge and marital status, tobacco use, birthplace, high school education or higher, or cervical cancer screening practices (Table 4).

Cervical Cancer Attitudes
Regarding participants' attitudes toward cervical cancer screening, 100% of participants agreed with the statement "I think it is important to get regular Pap tests." (Fig. 2). When asked about feeling confused when discussing cervical cancer with providers, the most common response (38.2%) was "neutral" on the Likert scale, followed by "agree" (29.4%). The responses indicating agreement were combined to create two categories of respondents; patients that felt confused when discussing cervical cancer (responded agree and strongly agree) versus patients who did not indicate feeling confused (responded neutral, disagree, and strongly disagree). A chi-squared test was run to determine if patients with inadequate versus adequate health literacy had differences in confusion levels, and a significant difference was found (p = 0.043). All participants indicated that they would follow up in the clinic if they received an abnormal Pap test result (Table 5).

Free Response Questions
Fourteen participants (41%) provided answers to the first free-response question, which asked patients to describe any barriers they have faced when seeking cervical cancer screening. Of these, seven (50%) responded in the negative, indicating that they perceived no barriers to their cervical cancer screening care. Four participants (29%) indicated that lack of health insurance is a barrier to their care. Two patients cited fear, either of the Pap test itself or of a positive result, as barriers to care. One patient stated that the embarrassment of being examined while naked is a barrier for her. Six patients (15%) responded to the second free response question, which inquired about any final comments that respondents would like to share. Of these, two emphasized that cervical cancer screening is an important part of preventative health. Two others indicated a desire for further information on the topic. One patient responded, "It's the most severe illness because there is no cure. I have seen many people die due to this." Figure 3 illustrates representative quotes in the form of a word cloud.

Discussion
In this quality improvement project, we assessed Spanish health literacy and cervical cancer knowledge in patients at the USF BRIDGE Healthcare Clinic, a student-run free clinic in Tampa, FL. We sought to understand our patients' ability to receive and process their health information and focused on cervical cancer screening due to the known increased risk of cervical cancer in Hispanic women [16]. For the first time, we have assessed the health literacy of BRIDGE patients; specifically, patients at risk for cervical cancer who receive care in Spanish. Of our sample of 34 patients, seven had inadequate health literacy as measured by the SAHL-S (21%). Additionally, we found that participants with inadequate health literacy scored lower on the cervical cancer knowledge questions, revealing a potential association between low Spanish health literacy and subsequent poorer understanding of cervical cancer ( Table 3).
As previous BRIDGE quality improvement projects have demonstrated, a high prevalence of adequate cervical cancer screening behaviors was found in our patient population [17]. However, we found no difference in rates of being up to date with Pap testing between BRIDGE patients with adequate versus inadequate Spanish health literacy. This indicates that the BRIDGE Clinic currently does a sufficient job at providing cervical cancer screening to patients regardless of health literacy. This is a notable difference from existing studies from similar population groups, where health literacy has been found to be associated with lower rates of cervical cancer screening [10,12]. We also found that 100% of our respondents felt it was important to get regular Pap tests and agreed they would follow up if they had an abnormal Pap test result, indicating that our patients are aware of the importance of cervical cancer screening.
On the other hand, it was noted that patients with inadequate health literacy were more likely to feel confused when discussing cervical cancer screening with their providers. This was further supported by project administrators' observations that patients with inadequate health literacy scores tended to take longer to fill out the questionnaire and needed things to be repeated more often. One patient indicated in the second free response question that "there is no cure" for cervical cancer, which illustrates a gap in understanding regarding the existence of curative surgical procedures for cervical cancer. These findings indicate that the BRIDGE Clinic can do a better job of communicating with patients to make them feel confident in their health care decisions regarding cervical cancer risk and screening. Other studies of similar population groups have come to this conclusion as well, and have pursued a variety of efforts to engineer creative solutions to their community's needs with limited success [10,12]. One method of addressing this gap in the BRIDGE Clinic's care would be to ensure that any educational materials and electronic communications provided are at an elementary school reading level or lower. We found that our respondents with an education level of middle school or higher were more likely to have adequate health literacy, so therefore the group of patients with the greatest need for customized health care guidance would be those with lower education levels. Further research is needed to determine why middle school may be a crucial time for the establishment of adequate health literacy.
Another strategy would be to consider training a select group of BRIDGE providers and interpreters on working with patients of lower health literacy. This would require a general screening of the health literacy of all BRIDGE patients to identify which patients need further support. Assessing this would also allow for patients with inadequate health literacy to potentially be assigned longer appointment times or scheduled on days with a lower patient volume. This would permit the specially trained providers enough time to utilize recommended techniques such as comprehension checks and repetition for patients with lower health literacy [18]. It would be helpful to know if such interventions improved patients' health literacy or reduced their perceived levels of confusion on certain health topics.
We do acknowledge some limitations of our project. First, this is a single-site survey at a student-run free clinic, and these results may not be generalizable to other Spanish-speaking patient populations. Although we had good variation in certain demographics such as birthplace, English proficiency, and education level, other demographics were more homogenous. Of note, the fact that our sample was entirely composed of underserved, low SES patients already seeking health care of their own accord may indicate that our population could have distinct health attitudes or behaviors, and potentially higher health literacy, than other underserved Spanish-speaking patient populations. Next, our response rate was somewhat low at 24.6% of eligible patients, as we were limited by clinic occurring only once weekly and interpreters being in short supply. Additionally, only 7 of the 34 participants were found to have inadequate health literacy, which led to relatively unequal comparison groups for statistical analysis. This was a particular concern when analyzing answers to the first knowledge question, where only two patients indicated having never heard of a Pap test. As such, chi-squared tests for this knowledge variable must be considered with this limitation in mind. Future studies of Spanish health literacy should consider sampling patients from multiple studentrun free clinics to access more patients for statistical analysis.
The results from this project reveal that Spanish health literacy may play a role in the growing disparity in cervical cancer morbidity and mortality among Hispanic women. Although the BRIDGE Clinic seems to do a sufficient job in providing cervical cancer screening to patients, there is a clear lack of understanding in patients with low Spanish health literacy as evidenced by their lower scores on the knowledge questionnaire. This could be an issue if our patients transfer care from BRIDGE to other clinics with less robust screening protocols or community resources. These findings also imply that patients of low health literacy may have poor comprehension of other aspects of their health beyond cervical cancer screening, such as vaccinations, colon cancer screening, and mammography. Going forward, we plan to implement some of the strategies discussed above to improve communication with BRIDGE patients of low Spanish health literacy. In doing so, we hope to improve our patients' comprehension of cervical cancer in addition to other vital aspects of their primary care.