Evaluation of Health Literacy in Plastic Surgery Using a Crowdsourced Patient Survey

Background:

Little is known about the levels of health literacy (HL) among plastic and reconstructive surgery (PRS) patients compared with the general population. This study aimed to characterize HL levels in patients interested in plastic surgery and identify potential risk factors associated with inadequate levels of HL among this population.

Methods:

Amazon’s Mechanical Turk was used to distribute a survey. The Chew’s Brief Health Literacy Screener was used to evaluate the level of HL. The cohort was divided into two groups: non-PRS and PRS groups. Four subgroups were created: cosmetic, noncosmetic, reconstructive, and nonreconstructive groups. A multivariable logistic regression model was constructed to assess associations between levels of HL and sociodemographic characteristics.

Results:

A total of 510 responses were analyzed in this study. Of those, 34% of participants belong to the PRS group and 66% to the non-PRS group. Inadequate levels of HL were evidenced in 52% and 50% of the participants in the non-PRS and PRS groups, respectively (P = 0.780). No difference in HL levels was found in the noncosmetic versus cosmetic groups (P = 0.783). A statistically significant difference in HL levels was evidenced between nonreconstructive versus reconstructive groups after holding other sociodemographic factors constant (0.29, OR; 95% CI, 0.15–0.58; P < 0.001).

Conclusions:

Inadequate levels of HL were present in almost half of the cohort, which highlights the importance of adequately assessing HL levels in all patients. It is of utmost importance to evaluate HL in clinical practice using evidence-based criteria to better inform and educate patients interested in plastic surgery.

Takeaways

Question: What is the level of health literacy (HL) in patients interested in plastic surgery and what are the associated risk factors with inadequate levels of HL?

Findings: Inadequate levels of HL were evidenced in 52% and 50% of the participants in the non-PRS and PRS groups, respectively. Statistically significant difference in HL levels was found between nonreconstructive and reconstructive groups.

Meaning: It is important to evaluate HL in clinical practice using evidence-based criteria to better inform and educate patients interested in plastic surgery.

INTRODUCTION

Health literacy (HL) plays a critical role when patients are faced with decisions that will ultimately impact their health.^1,2 Proper assessment of HL is of utmost importance in order to provide appropriate and specific patient education and, ultimately, care. As patients obtain and assimilate information related to their care, they progressively feel empowered and are increasingly willing to actively partake in their medical decisions.^1,2 Furthermore, studies have shown that adequate levels of HL positively impact patients’ participation and compliance in their management, as well as improve patient-reported and clinical outcomes.^3–5 Understanding levels of HL in our patient population remains a challenge, and further insight could help in the process of patient education and counseling and ultimately improve overall patient outcomes. Identifying a patient’s HL level is vital to being able to communicate to patients before any intervention, surgical or medical.⁶

It is well recognized that low or inadequate levels of HL have been associated with poor health outcomes, increased rates of chronic disease, and higher mortality rates.^7,8 In addition, these patients have increased rates of hospitalizations, emergency department visits, and poor medication adherence.^9–12 These issues disproportionately affect those with disabilities, low socioeconomic status, racial/ethnic minorities, non-English proficient patients, those with lower levels of education, and older patients.^9,13,14 In 2003, a study showed that 36% of American adults had basic or below basic levels of HL in the National Assessment of Adult Literacy.¹⁵

Studies establishing the levels of HL among plastic and reconstructive surgery (PRS) patients compared to the overall population have been lacking so far. Assessing the levels of HL may help shed light on ways to empower patients to make informed health care decisions, build a solid patient-physician dynamic, and improve health care outcomes. Therefore, this study aimed to determine the levels of HL in the PRS population, as well as to assess potential risk factors associated with inadequate levels of HL among this population.

METHODS

Study Design and Setting

This is an IRB-approved (2021P000781) cross-sectional study to assess the level of HL among the US population and participants interested in PRS. A survey was distributed through Amazon Mechanical Turk (MTurk) (Amazon Mechanical Turk, Seattle, Wash.) between October 20, 2021, and November 11, 2021. The double-encrypted Research Electronic Data Capture was used to create and design the survey. Three sections were included in the survey: (1) sociodemographic characteristics, (2) plastic surgery interest questions, and (3) the Chew’s Brief Health Literacy Screener. (See questionnaire, Supplemental Digital Content 1, which shows the questions asked in this study, http://links.lww.com/PRSGO/C390.)

In order to target the US population, we distributed the survey through MTurk to subjects with the following characteristics: Human Intelligent Task (HIT) Approval Rate (%) for all requesters’ HITs greater than 95 and located in the United States. As defined by MTurk “a HIT represents a single, self-contained, virtual task that a worker can work on, submit an answer to, and collect a reward for completing.”¹⁶ Plastic surgery interest questions were used to broadly divide the cohort into two groups: (1) a non-PRS group including participants with no background or interest in PRS; and (2) a PRS group including participants interested in or who previously had any PRS. Moreover, patients were inquired about their surgical history regarding reconstructive or aesthetic procedures and their interest in aesthetic surgery. Based on the previously mentioned, the cohort was further divided into four subgroups: (1) cosmetic group including participants interested in cosmetic surgery, (2) noncosmetic group including participants not interested in cosmetic surgery; (3) reconstructive group including participants who had previously undergone reconstructive surgery, and (4) nonreconstructive group including participants who have not had any reconstructive surgery (Fig. 1). It is noteworthy that we inquired patients about what type of surgical intervention was performed to corroborate cosmetic and reconstructive procedures.

Fig. 1.

Flow diagram of the cohort division based on participants' answers.

The level of HL among the responders was evaluated using Chew’s Brief Health Literacy Screener, a validated screening questionnaire.¹⁷ This screening tool is based on three single-item questions: (1) “How often do you have someone help you read hospital materials?” (2) “How often do you have problems learning about your medical condition because of difficulty understanding written information?” and (3) “How confident are you filling out medical forms yourself?” Each question is structured to be answered with a five-point Likert scale (0–4). (See questionnaire, Supplemental Digital Content 1, which shows the questions asked in this study, http://links.lww.com/PRSGO/C390.) The final score is the sum of all the data points gathered from each question. Based on the optimized sensitivity and specificity reported in previous studies, a cutoff of 6 was determined.^17,18 Therefore, participants with scores greater than or equal to 6 were categorized as having inadequate HL.^17,18 Conversely, participants with scores less than 6 were categorized as having adequate HL.^17,18

Inclusion criteria were all participants who currently reside in the USA and are 18 years or older. Participants were provided a prospective agreement for the survey, and completion was considered consent to use their responses. Respondents were compensated $0.25.

Statistical Analysis

For descriptive analysis, frequencies and percentages were used for categorical variables, and means and standard deviations were used for continuous variables as normal distribution was assumed following the central limit theorem. Differences between groups were evaluated using the unpaired t test and Fisher exact test for continuous and categorical variables, respectively. Moreover, a multivariable logistic regression model was constructed to assess associations between levels of HL and sociodemographic characteristics. Statistical significance was set at an alpha value of 0.05. Stata/IC 16.1 (StataCorp LLC, College Station, Tex.) was used to conduct statistical analysis.

RESULTS

Participants Characteristics

A total of 510 participants were included. Of those, 175 (34%) participants comprised the PRS group and 335 (66%) the non-PRS group. The mean age was 39.1 SD 11.8 and 38.2 SD 13.1 for the PRS and non-PRS groups, respectively. The non-PRS group had more men (59%) than the PRS group (47%). Most of the cohort self-identified as White (84% in the PRS group and 82% in the non-PRS group) and non-Hispanic (81% in the non-PRS group and 78% in the PRS group). Both groups had similar percentages of highly educated participants or with at least some college education (96% for both the non-PRS and PRS groups). No statistically significant difference was evidenced between sociodemographic characteristics between groups, except for gender (<0.011) (Table 1).

Table 1.

Population Characteristics

	Non-PRS	PRS	P
Total, n (%)	335 (66)	175 (34)
Age, mean (SD)	38.2 (13.1)	39.1 (11.8)	0.4316
Gender, n (%)
Male	196 (59)	82 (47)	0.011*
Female	139 (41)	92 (53)
Nonbinary	0 (0)	1 (0.6)
Other	0	0
Race, n (%)
White	276 (82)	147 (84)	0.254
Black or African American	29 (9)	14 (8)
Native American	3 (1)	5 (3)
Asian	23 (7)	6 (4)
Native Hawaiian	0	0
Other	4 (1)	2 (1)
Ethnicity, n (%)
Hispanic or Latinx	64 (19)	38 (22)	0.486
Non-Hispanic or Latinx	271 (81)	137 (78)
English proficiency, n (%)	328 (98)	170 (100)	0.101
Education, n (%)
Low education	15 (5)	7 (4)	1.000
High education	320 (96)	168 (96)
HL level
Adequate	161 (48)	87 (50)	0.780
Inadequate	174 (52)	88 (50)

%, percentage; n, frequency.

^*Statistically significant (p < 0.05).

A total of 159 (31%) and 54 (11%) participants were included in the cosmetic and reconstructive groups, respectively. The mean age for the cosmetic group was 38.9 SD 11.9 and for the reconstructive group was 38.2 SD 11.9. Women were predominant in the cosmetic group (57%), and men were predominant in the reconstructive group (61%). The majority were White (84% cosmetic group and 83% reconstructive group) and non-Hispanic (79% cosmetic group and 67% reconstructive group). Moreover, in the cosmetic and reconstructive groups, most of the participants were highly educated in 96% and 98% of the cases, respectively. A statistically significant difference was present between race (P = 0.033) and ethnicity (P = 0.031) between the reconstructive and nonreconstructive groups. On the other hand, a significant difference was evidenced in gender between cosmetic and noncosmetic groups (P < 0.001) (Table 2).

Table 2.

Cosmetic and Reconstructive Groups

	Cosmetic Group	P *	Reconstructive Group	P *
Total, n (%)	159 (31)		54 (11)
Age, mean (SD)	38.9 (11.9)	0.6393	38.3 (11.9)	0.8975
Gender, n (%)
Male	68 (43)	<0.001†	33 (61)	0.451
Female	90 (57)		21 (39)
Nonbinary	1 (1)		0
Other	0		0
Race, n (%)
White	133 (84)	0.565	44 (83)	0.033†
Black or African American	13 (8)		6 (11)
Native American	4 (3)		3 (6)
Asian	6 (4)		0
Native Hawaiian	0		0
Other	2 (1)		0
Ethnicity, n (%)
Hispanic or Latinx	33 (21)	0.811	17 (32)	0.031†
Non-Hispanic or Latinx	126 (79)		37 (67)
English proficiency, n (%)	155 (100)	0.106	50 (100)	1.000
Education, n (%)
Low education	6 (4)	0.816	1 (2)	0.496
High education	153 (96)		53 (98)
HL level
Adequate	79 (50)	0.775	41 (76)	<0.001†
Inadequate	80 (50)		13 (24)

^*Compared with their counterpart (no-cosmetic versus no-reconstruction).

^†Statistically significant (p < 0.05).

Health Literacy

Inadequate levels of HL were found in 52% and 50% of the participants in the non-PRS and PRS groups, respectively, representing no statistically significant difference between HL levels among the two groups (P = 0.780) (Table 3). No differences in HL levels were found in the noncosmetic versus cosmetic groups (P = 0.775) (Table 4). A statistically significant difference in HL levels was evidenced between the nonreconstructive versus reconstructive groups [0.29, OR; 95% CI (1.15–0.58); P < 0.001] (Table 5).

Table 3.

Multivariable Analysis of HL Levels based on Sociodemographic Characteristics: Patients Who Had Interest in Plastic Surgery versus Not

Variable	HL		P
	OR	95% CI
Non-PRS versus PRS
Non-PRS	Reference
PRS	0.96	0.66–1.41	0.850
Age	1.03	1.01–1.04	<0.001*
Ethnicity
Non-Hispanic	Reference
Hispanic	0.59	0.37–0.94	0.026*
Education
Low	Reference
High	0.45	0. 17–1.21	0.115
No-English proficiency	0.39	0.07–2.08	0.268
Race
White	Reference
Black or African American	1.18	0.61–2.26	0.625
Asian	1.49	0.67–3.29	0.326
Others	1.89	0.58–6.13	0.291

^*Statistically significant (p < 0.05).

Table 4.

Multivariable Analysis of HL Levels based on Sociodemographic Characteristics: Patients Who Had Interest in Cosmetic versus Not

Variable	HL		P
	OR	95% CI
Noncosmetic versus cosmetic
Noncosmetic	Reference
Cosmetic	0.96	0.65–1.42	0.836
Age	1.03	1.01–1.04	<0.001*
Ethnicity
Non-Hispanic	Reference
Hispanic	0.59	0.37–0.94	0.029*
Education
Low	Reference
High	0.45	0.17–1.21	0.116
No-English Proficiency	0.39	0.07–2.08	0.268
Race
White	Reference
Black or African American	1.18	0.61–2.27	0.624
Asian	1.49	0.67–3.29	0.324
Others	1.89	0.58–6.14	0.289

^*Statistically significant (p < 0.05).

Table 5.

Multivariable Analysis of HL Levels based on Sociodemographic Characteristics: Patients Who Had Reconstructive Surgery versus Not

Variable	HL		P
	OR	95% CI
Nonreconstructive versus reconstructive
Nonreconstructive	Reference
Reconstructive	0.29	0.15–0.58	<0.001*
Age	1.03	1.01–1.04	<0.001*
Ethnicity
Non-Hispanic	Reference
Hispanic	0.60	0.38–0.97	0.037*
Education
Low	Reference
High	0.48	0.18–1.30	0.149
No-English proficiency	0.35	0.07–1.86	0.218
Race
White	Reference
Black or African American	1.20	0.61–2.33	0.599
Asian	1.34	0.60–2.94	0.481
Others	2.01	0.58–6.91	0.268

^*Statistically significant (p < 0.05).

Also, increasing age was associated with higher odds of inadequate levels of HL, while Hispanic ethnicity was associated with lower odds of inadequate levels of HL after holding other sociodemographic factors constant (3">Tables 3–5). However, education, English proficiency, and race were not found to be associated with levels of HL. Moreover, when conducting the multivariable logistic model in the PRS group, age [1.03, OR; 95% CI (1.00–1.06); P = 0.045] and Hispanic ethnicity [0.36, OR; 95% CI (0.16–0.84); P = 0.018] remained as risk factors for inadequate levels of HL.

DISCUSSION

Plastic surgery patients have similar, inadequate levels of HL compared to the overall population in the United States. This study represents the first of its kind in assessing levels of HL among the PRS population, as well as its difference with the overall, general population, using a crowdsourcing interface platform. Almost half of the cohort was found to have low levels of HL. Moreover, when compared to the general population, the PRS interest group had similar levels of HL. Furthermore, age and ethnicity were found to be factors associated with the level of HL among the PRS group. Such findings bring up the importance of adequately assessing HL levels among PRS patients in both clinic and hospital settings as well as to properly identify sociodemographic backgrounds to provide patient-specific and culturally appropriate care.

Previous reports suggest that only 12% of US adults had proficient HL levels, and over a third of US adults had difficulty with common health tasks (eg, following directions on a prescription drug label).¹⁹ The National Assessment of Adult Literacy has shown that individuals with low HL are more likely to be 65 years or older, male, Black or Hispanic, have English as a second language, have lower education level, live at or below the poverty line, and rate their overall health as poor.^15,20–22 Interestingly, our findings suggest that participants who self-identified as Hispanic had lower odds of having inadequate levels of HL; however, we believe this might be secondary to two reasons: (1) participants who self-identified as Hispanic in our cohort were highly educated, and (2) this population must be technologically literate in order to use MTurk. Moreover, Baker et al²³ found that patients with inadequate HL had more than twice the risk of being hospitalized in a two-year period. Additionally, a previous study showed that low HL is associated with nonadherence to preoperative medication instructions and ultimately avoidable surgery cancelations and postponements, wasted operating room time, and additional hospital expenses.²⁴ This is a compounded issue since patients may also struggle with nonwritten forms of communication, making the conversations between physicians and patients challenging, which could also reflect on other areas of care such as the informed consent process.

Although trends in the PRS group are similar to the levels of HL within the US population, reconstructive surgery patients demonstrated higher levels of HL compared to the general population. Several reasons could explain the previous finding. One potential explanation might be the prolonged longitudinal and multidisciplinary care that is part of their journey. Usually, patients who require reconstructive surgery have diseases that necessitate long-term health care exposure (eg, cancers), which could contribute to increased levels of HL in these patients. Likewise, the complexity of reconstructive surgery in and of itself requires extensive patient education, further exposing these patients to the health care system as part of the in-depth procedural and postoperative care discussion. Finally, reconstructive surgery usually requires frequent follow-up visits, as well as surgical staging and possibly revisional procedures, which further expose the patients to the health care environment.

To our knowledge this is the first time, HL levels were assessed in the cosmetic surgery population. Our findings show that there were no significant differences in HL levels between the cosmetic surgery and noncosmetic surgery subgroups. Additionally, we found that half of the cosmetic group had inadequate levels of HL, which shows that this is a common issue.

Solutions to potential gaps in HL involve improving physician communication and patient education using easily accessible and simplified informational material and combining it with other educational methods, which include the teach-back method. The teach-back method has been previously studied as a method of verifying patient understanding and has been used in the context of overcoming the hurdle of health information exchange.²⁵ This method consists of checking patients’ understanding by asking them to verbalize in their own words what they assimilated about their health and the action items that follow.²⁶ Both the Agency for Research and Quality and the Institute for Healthcare Improvement have recommended this technique due to its perceived efficacy in detecting miscommunication or misunderstanding between providers and patients.²⁵ Plastic surgeons in particular underutilize this technique.² In fact, studies have shown that plastic surgeons used the following methods when counseling patients: lay terminology in 94% of the cases, pictures/diagrams in 84.6%, and teach-back strategies in 8.1%.^2,27 Therefore, we believe that further discussion and physician education have the potential to increase rates of utilization.

This study should be evaluated within the scope of its limitations. Participants using this crowdsourcing internet marketplace platform might not be representative of the general US population. Potential reasons for nonrepresentative results include but are not limited to the following: participants are financially remunerated to answer surveys and must be both literate and technologically savvy. However, studies using Mturk have previously demonstrated it to be useful to the scientific literature despite these shortcomings.^28–32 Also, unmeasured confounders could potentially affect the current analysis. Moreover, the Chew’s Brief Health Literacy Screener is a screening tool that is subject to a false negative result. Therefore, standard questionnaires, such as the Short Test of Functional Health Literacy in Adults and the Rapid Estimate of Adult Literacy in Medicine, might be needed to corroborate our findings.

Future patient-based survey studies should be conducted to correlate this study’s findings. Likewise, studies assessing the impact of levels of HL on clinical and patient-reported outcomes are needed, as well as studies evaluating the efficacy of evidence-based communication methods, such as the teach-back method, among the PRS population.

CONCLUSIONS

Inadequate levels of HL were present in almost half of the cohort studied, which highlights the importance of adequately assessing HL levels among PRS patients in both clinics and the hospital setting. Interestingly, our study results suggest that reconstructive patients had lower odds of having inadequate HL levels compared to patients who had not undergone any reconstructive procedure. It is of utmost importance to enhance patient HL in the clinical practice and hospital setting by using evidence-based methods, such as the teach-back method, to better serve the patient populations within PRS with inadequate HL and minimize potential poor health outcomes.

Supplementary Material

gox-11-e4803-s001.pdf ^{(102.9KB, pdf)}