|
 
 |
| ORIGINAL ARTICLE |
|
| Year : 2022 | Volume
: 6
| Issue : 4 | Page : 267-275 |
|
Medical maximizing-minimizing preferences and health beliefs associated with emergency department patients' intentions to take a cardiac stress test after receiving information about testing
Andrew J Foy1, Ashley Bucher2, Lauren J Van Scoy3, Laura D Scherer4
1 Department of Medicine, Penn State M.S. Hershey Medical Center and College of Medicine, Penn State University Heart and Vascular Institute; Department of Public Health Sciences, Penn State M.S. Hershey Medical Center and College of Medicine, Hershey, Pennsylvania, USA
2 Department of Medicine, Penn State M.S. Hershey Medical Center and College of Medicine, Penn State University Heart and Vascular Institute, Hershey, Pennsylvania, USA
3 Department of Public Health Sciences, Penn State M.S. Hershey Medical Center and College of Medicine; Department of Medicine, Division of Pulmonary, Allergy and Critical Care Medicine, Penn State M.S. Hershey Medical Center and College of Medicine, Hershey, Pennsylvania, USA
4 Department of Medicine, Adult and Child Consortium for Health Outcomes Research and Delivery Science, University of Colorado, Colorado, USA
| Date of Submission | 09-Aug-2021 |
| Date of Acceptance | 06-Apr-2022 |
| Date of Web Publication | 08-Nov-2022 |
Correspondence Address:
Andrew J Foy
MD, 500 University Dr Hershey, PA 17033
USA
 Source of Support: None, Conflict of Interest: None  | Check |
DOI: 10.4103/hm.hm_48_21
Purpose: The Medical Maximizer-Minimizer Scale (MMS) has been validated to predict preferences for health-care service use on hypothetical vignettes in nonclinical cohorts. Using mixed methods, we sought to determine whether it would predict preferences for cardiac stress testing in a cohort of emergency department (ED) patients with low-risk chest pain within the context of the Health Belief Model (HBM). Design: Patients who met the definition for low-risk chest pain and who were eligible to take a cardiac stress test before being discharged from the hospital were recruited to participate. Each participant provided demographic information and completed the MMS-10 paper-and-pencil scale. They then watched a 7-min informational video on an iPad tablet that provided information about the condition of “low-risk chest pain” and the probabilities of results and outcomes following a cardiac stress test. After the video, participants answered a one-question survey on their intention-to-take (ITT) a cardiac stress test or not and were then interviewed about factors that influenced their decision-making. Interviews were interpreted using a HBM lens. Results: Sixty participants were enrolled in the study who were between the ages of 29 and 80 years with a mean age of 53 (± 10.8); 58% were women and 90% were white. The mean MMS score was 4.6 (± 0.8) and ranged from 2.6 to 6.8. Minimizers accounted for 25% (n = 15) of the cohort while maximizers accounted for 75% (n = 45). MMS scores followed a normal distribution and were found to be mildly correlated with ITT scores (r = 0.25; P = 0.051). The mean ITT scores for individuals with MMS scores in the 1st and 4th quartiles were 3.9 ± 2.2 and 5.9 ± 1.7, respectively. After watching the informational video, individuals' perceptions related to the: (1) low threat posed by the condition, (2) low utility (low benefits + significant barriers) of taking a cardiac stress test, and (3) high benefits of taking a cardiac stress test were all strongly associated with ITT scores in a directional manner. No direct connection was found between minimizer-maximizer preferences and health beliefs after watching the informational video. This may have been due to sample size and underrepresentation of minimizers in the cohort. Conclusions: MMS and health beliefs predicted preferences for cardiac stress testing in ED patients with low-risk chest pain after viewing an informational video on the topic. However, we did not find direct evidence that the relationship between MMS and decision-making is mediated through the formation of perceptions of threat and utility consistent with the HBM. More research is needed to establish this connection and understand how framing of information in the health-care space may interact with stable personality traits to influence decision-making.
Keywords: Chest pain, Health Belief Model, medical decision-making, risk communication, stress testing
How to cite this article:
Foy AJ, Bucher A, Van Scoy LJ, Scherer LD. Medical maximizing-minimizing preferences and health beliefs associated with emergency department patients' intentions to take a cardiac stress test after receiving information about testing. Heart Mind 2022;6:267-75 |
How to cite this URL:
Foy AJ, Bucher A, Van Scoy LJ, Scherer LD. Medical maximizing-minimizing preferences and health beliefs associated with emergency department patients' intentions to take a cardiac stress test after receiving information about testing. Heart Mind [serial online] 2022 [cited 2023 May 29];6:267-75. Available from: http://www.heartmindjournal.org/text.asp?2022/6/4/267/360608 |
| Introduction | |  |
Chest pain represents the second leading cause of adult emergency department (ED) visits, and it has been repeatedly demonstrated that those who rule out for a heart attack (the majority) have a low risk of dying or having a heart attack in the near future and that taking a cardiac stress test before hospital discharge is not associated with a lower risk of future events.[1],[2],[3],[4],[5],[6],[7] Research shows tremendous variation in inhospital stress testing rates for patients meeting the definition of “low-risk chest pain” which attests to the significant equipoise that exists.[8] Cardiac stress testing in this circumstance is regarded as a “preference-sensitive test,” and researchers have found that patients randomized to shared decision-making using a decision support tool are less likely to choose to undergo cardiac stress testing compared to those randomized to standard care.[9] However, research has not yet sought to explore how psychological traits may influence decision-making in this or other similar clinical circumstances.
The Medical Maximizer-Minimizer Scale
The Medical Maximizer-Minimizer Scale (MMS) assesses the degree to which individuals prefer to seek or avoid medical care. The notion of medical maximizing and minimizing is based on the psychological theory of commission and omission bias, which proposes that individuals have a general preference for action or inaction and that these preferences extend to health-care-related decisions. Maximizers prefer to seek care whereas minimizers prefer to avoid it.
Initial validation of the 10-item MMS (MMS 10) showed that it was distinct from hypochondriasis, belief that doctors do harm, health-care access, and health status, and predicted self-reported health-care utilization and a variety of treatment preferences independent of age, sex, race, education, occupation in medical field or not, insurance status, and health status.[10] In a subsequent study, MMS-10 was found to be independently associated with online survey respondents' preferences for prostate-specific antigen (PSA) screening.[11] Initially, the majority of men (76.8%) wanted PSA screening; however, after receiving information about the potential harms of PSA screening, interest declined to 58.8%, and men who changed their preferences from yes to no were significantly more likely to be minimizers than maximizers. After being informed, approximately 80% of strong maximizers (MMS 90th percentile) wanted PSA screening compared to only 25% of strong minimizers (MMS 10th percentile).
In another study of online survey respondents, MMS-10 was found to predict preferences for high- and low-benefit care.[12] Among the 785 United States (US) adults, responding to 18 hypothetical scenarios (8 considered high-benefit and 10 considered low-benefit care), strong maximizers were significantly more likely to prefer receiving high-benefit care whereas strong minimizers were significantly more likely to want to avoid low-benefit care. The MMS-10 independently explained 11% of variation in preferences for high-benefit care (e.g., taking medication for established high blood pressure), and 29% of variation in preferences for low-benefit care (e.g., a full-body computed tomography scan in healthy adults).
To date, the medical maximizer-minimizer score has been demonstrated to be a consistent predictor of health-care preferences in hypothetical vignettes. However, it has yet to be tested in a real-life situation where other factors and emotions may dampen or extinguish the relationship.
The Health Belief Model
The Health Belief Model (HBM) was developed in the 1950s by social psychologists at the U.S. Public Health Service and is one of the most widely used theories in health behavior research.[13] The HBM posits that health-related action depends on[1] an individual's perception of threat related to an illness and[2] his or her perception that the action in question will prevent or fix it (i.e., utility).[13] Specifically, the HBM consists of the following dimensions: perceived susceptibility and perceived severity which collectively compose the domain of threat, and perceived benefits and perceived barriers which collectively compose the domain of utility.[13] To trigger the decision-making process, a stimulus or action cue is needed, which can be either internal (i.e., symptoms) or external (e.g., health awareness campaign).[13]
When represented graphically, HBM dimensions are typically presented as items most proximally related to health action(s) whereas individual characteristics (e.g., sex, age, and socioeconomic status) and psychological traits are presented as upstream factors that mediate the formation of health beliefs [Figure 1]. As depicted, a health-related action depends on the interplay of perceptions of threat and utility that each individual assigns various weights to on a subjective basis.{Figure 1}
In a review of 46 HBM-related investigations, Janz and Becker found substantial empirical support for the HBM.[13] Most studies tested either a preventive health behavior (n = 24) or a sick-role behavior (n = 19), and there were more retrospective (n = 28) than prospective (n = 18) designs.[13] “Perceived barriers” was found to be the most powerful HBM dimension across the various study designs and behaviors; however, empirical support was found for all HBM dimensions.[13]
The interplay between Maximizer-Minimizer Scale and Health Belief Model dimensions in a real-life decision-making scenario
The present research study used a convergent mixed-methods design that involved collection of both quantitative and qualitative data to assess how and why individuals under consideration for cardiac stress testing in the emergency room make decisions about further workup after receiving information about their risks of a bad outcome as well as the potential benefits and risks of taking a cardiac stress test. To do so, we collected quantitative data to investigate participants' MMS orientation and used qualitative interviews to explore similar constructs related to health beliefs. Data were integrated to examine the interplay between the MMS and HBM dimensions.
We hypothesized that MMS would correlate with intention-to-take (ITT) scores such that higher MMS scores, reflecting maximizing tendencies, would associate with higher ITT scores. We also hypothesized that HBM constructs that included individual's perceptions related to the susceptibility and severity of their condition along with perceptions of the benefits and barriers to stress testing, after watching an informational video, would correlate with decision-making. Furthermore, we hypothesized that MMS would correlate with HBM constructs in a directional manner such that people with a maximizing tendency would perceive higher threat from the condition of chest pain and higher utility from having a stress test performed and people with a minimizing tendency would perceive lower threat and lower utility of stress testing.
| Methods | | |
Participants
Participants were adults 18 years of age and older who presented to the emergency medicine (EM) department of a single academic medical center with a chief complaint of chest pain who ruled out for an acute myocardial infarction (MI) by having at least one electrocardiogram (ECG) with no evidence of acute myocardial ischemia, based on the interpretation of their treating EM provider, and two troponin tests, drawn at least 4 h apart with a value that was <99% upper reference limit of the test. Patients meeting the above criteria fulfill the definition of “low-risk chest pain.”[14] Patients were excluded if they did not meet the above specifications, if they had another condition requiring hospital admission, or if they did not speak English.
Based on institutional protocol, patients meeting the above criteria, without another clinical condition requiring admission to the hospital, are offered the option to take an inpatient cardiac stress test with further management based on the test's results or to be discharged home with outpatient follow-up arranged with either their primary care physician or a cardiologist within 1 week.
Eligible patients were identified by the research study coordinator using the tracking board in the EM department that is embedded within the electronic medical record system of the institution, or the research study coordinator was contacted directly by EM providers to provide notification of a potentially eligible participant. Regardless of who initiated contact, the EM physician providing direct care for the potentially eligible patient had to confirm patient eligibility and provide verbal consent to allow the research assistant to approach the patient for recruitment. As criteria for enrollment, a final decision regarding inpatient stress testing or going home without stress testing could not have been previously decided between the patient and their EM provider.
The study was approved by the institutional review board. All patients provided written informed consent to participate in the study, which was determined to present “minimal risk” to patients. We determined we would need to recruit at least 47 participants to detect a correlation of r = 0.4 based on a two-sided alpha of 0.05 and beta of 0.20. It was predetermined that if the recruitment target was reached before December 31, 2019, that we would continue to recruit patients in an effort to enrich our qualitative analysis, which did not have a recruitment target.
Study procedures and measures
After agreeing to participate in the study, individuals completed a demographics questionnaire and the MMS-10, with each item scored on 7-point Likert scale that asks participants to rate how much they agree or disagree with each statement [1 = disagree, 7 = strongly agree; [Figure 1]].[1] The scores for each question are averaged for a final score ranging from 1 to 7.
Participants then watched a 7-min informational video that was presented on an iPad tablet. The video was not intended to be used as an “intervention” in this study but instead as a way to provide uniform information to each participant. The video is divided into six sections that explain: (1) the condition of low-risk chest pain and risk of future events, (2) treatment options (i.e., undergoing cardiac stress test before discharge versus going home without stress testing, (3) the probabilities of positive and negative stress test results, (4) the probabilities of true-positive and false-positive results and an explanation of what these terms mean, (5) risks associated with and explanation of subsequent invasive testing and treatment (i.e., cardiac catheterization and revascularization), and (6) explanation of the 30-day risk of MI for patients deciding to take a stress test or not. Participants could watch the video more than once, if requested, and could ask for clarification of information provided in the video from the research assistant who could provide it. However, study assistants were explicitly instructed not to provide medical advice, outside of clarifying any information explicitly presented in the video, or to provide personal opinions (e.g., if a patient asked “What would you do if you were me?”).
The video was created by the lead investigator (AF) in conjunction with providers from the EM department. Before dissemination, it was reviewed by several patient focus groups and presented to all EM staff in a series of meetings. Patient focus groups and EM providers had the opportunity to suggest modifications, and several versions were created before final departmental approval. While it was created for the purpose of conducting this study, it was available to be used by any EM providers caring for patients who felt it was suitable. It has remained available for use after this study was completed.
The video was organized into six sections. Sections 1 and 2 provide information about chest pain and features that typically distinguish cardiac from noncardiac chest pain. Information on ECG testing and troponin is provided, explaining that these tests did not show evidence of a heart attack and that the patient's risk for experiencing a heart attack or dying in the next 30 days is <1% (1/100).
Sections 3 and 4 present patients with a “cardiac testing theater” and information on the probability of outcomes for 1000 theoretical patients who undergo stress testing in this circumstance. The probability of a positive result is presented as 50/1000 (5%). It is explained that patients with a positive result would be encouraged to have a cardiac catheterization. Descriptive and animated detail is provided on how a cardiac catheterization is performed and information is provided on the risk of complications from the procedure (i.e., major bleeding, heart attack, or stroke) as 1/50 (2%). The probability of a false-positive result is presented as 30/50 (60%) and compared descriptively and visually to a true-positive result, which is presented as 20/50 (40%). Explanations are provided on the difference between what it means to have a false-positive versus a true-positive test. For patients with true-positive results, it is explained that they would likely have a cardiac stent placed. Descriptive and animated detail is provided on how cardiac stenting is performed. The risk of complications from the procedure is presented as 1/20 (5%). The chance of benefitting from the procedure by having a heart attack prevented is presented as 5/20 (25%).
Sections 5 and 6 summarize information on a patient's risk of experiencing a heart attack or dying within 30 days based on whether they have a stress test (5/1,000) or not (10/1,000). Reassurance is provided that most patients would not have any abnormalities detected or adverse events regardless of whether they choose a stress test (945/1,000) or not (990/1,000).
After participants viewed the video, and were satisfied that they did not have any further questions about the information provided, they answered a single question (on a 7-point Likert scale) regarding the likelihood that they will choose to undergo cardiac stress testing as opposed to go home with outpatient follow-up (1 = very unlikely, 7 = very likely).
A semi-structured interview was then conducted to explore participants' decision-making process. The interviewer asked the participant to explain what information (if any) in the video influenced their thinking about whether to stay for a stress test or go home without stress testing. Questions then probed whether and how specific information provided in the video on overall risk of having a heart attack and the probabilities of stress test results including false-positive and true-positive findings as well as future procedures and potential complications (e.g., cardiac catheterization and stenting) influenced their thought process. Following the interview, patients were informed that their participation in the research study was over and that they would see their primary EM provider to finalize their disposition regarding staying for a stress test or going home. They were told that information provided and attained in the survey questions and interview would not be provided to their primary EM provider.
Transcript analysis and data transformation
All interviews were transcribed verbatim. The HBM was used to guide the analytic process. Data were analyzed using an inductive, descriptive content analysis approach followed by a data transformation that converted the qualitative concepts into categorical variables. To develop the preliminary codebook, coauthors AF and AB independently reviewed approximately 10% of the data (n = 10 transcripts). A list of broad categories and subcategories that emerged were assembled into a preliminary coding scheme. Within each category, codes with detailed definitions were created through group discussion. A final coding scheme was created after collapsing and refining the categories. Next, AF and AB used the final coding scheme to code the remaining 50 transcripts using the constant comparative method using NVivo 11 for Windows.[15] Coding conflicts were resolved through group discussion [Table 1].{Table 1}
After all interviews were coded, the qualitative data were transformed into dichotomous variables (yes/no) based on whether patients did or did not endorse each of the following categories or subcategories: high perceived threat, low perceived threat, high perceived benefit, low perceived benefit, significant perceived barrier, and strong internal stimulus (action cue) [Table 1].
Statistical analysis
All analyses were performed using SAS, version 9.4 (SAS Institute, Cary, NC, USA), and all variables were summarized before analysis using descriptive statistics. MMS and ITT scores were correlated using Pearson correlation coefficient. Student's t-tests and Fisher's exact tests were used for group comparisons where appropriate.
| Results | | |
Participant characteristics and quantitative data
Sixty participants were enrolled in the study who were between the ages of 29 and 80 years with a mean age of 53 (±10.8); 58% were women and 88% were white [Table 2]. The mean MMS score was 4.6 (±0.8) and ranged from 2.6 to 6.8. Minimizers (MMS score ≤4) accounted for 25% (n = 15) of the cohort while maximizers (MMS score >4) accounted for 75% (n = 45). MMS scores followed a normal distribution with a skewness of 0.08 [Figure 2]. There were no statistically significant associations between age (r = 0.19; P = 0.15) or sex (women [4.6 ± 0.7] vs. men [4.7 ± 1.0]; P = 0.53) and MMS scores.{Figure 2}{Table 2}
The mean ITT score of participants was 4.8 (±2.3) and ranged from 1 to 7. MMS score was mildly associated with ITT (r = 0.25) and nearly reached statistical significance (P = 0.051). When the cohort is divided on the basis of MMS score quartiles, individuals in the 1st quartile (MMS range: 2.6–4.0) had the lowest mean ITT score (3.9 ± 2.2). Participants in the 2nd quartile (MMS range: 4.1–4.6) had a mean ITT score of 5.1 ± 2.3. Individuals in the 3rd quartile (MMS range: 4.6–5.3) had a mean ITT score of 5.6 ± 2.1. Participants in the 4th quartile (MMS range: 5.3–6.8) had the highest average ITT score (5.9 ± 1.7).
Content analysis and group comparisons using transformed qualitative data
High perceived threat
Fifteen participants reported being strongly influenced by the perception that their condition posed a high threat to them [Table 3]. These involved perceptions related to their high personal susceptibility for experiencing a bad outcome (e.g., knowing someone with a similar story who ended up having a heart attack) as well as perceptions about the high potential severity of a heart attack if it were to occur [Table 3]. Individuals who endorsed a high threat construct (n = 15) had higher ITT scores compared to those who did not (n = 45), but the difference did not reach statistical significance (5.5 vs. 4.5; P = 0.15).{Table 3}
Low perceived threat
Twenty participants reported being strongly influenced by the perception that their condition did not pose a high threat to them [Table 3]. These encompassed statements related to their perceptions of having a low personal susceptibility for experiencing a serious outcome (e.g., the idea of doing anything with potential risks knowing the probability of a serious outcome, like having a heart attack, was so low) as well as their perceptions related to the low potential severity of the condition which led them to seek care [Table 3]. Individuals who endorsed a low threat domain had significantly lower ITT scores compared to those who did not (3.1 vs. 5.6; P < 0.01).
High perceived benefit (i.e., utility)
Twenty-three participants reported being strongly influenced by the perception that taking a cardiac stress test would be beneficial [Table 3]. These encompassed themes related to finding an answer, lowering the future risk of a heart attack, and the most common, to provide reassurance. Individuals who endorsed a high benefit domain had significantly higher ITT scores compared to those who did not (6.3 vs. 3.8; P < 0.01).
Low perceived benefit and/or significant barriers (i.e., utility)
Nineteen participants reported being strongly influenced by the perception that taking a cardiac stress test would not be beneficial to them or that there were significant barriers to doing so (we grouped these constructs together into the unifying concept of “low utility”) [Table 3]. These encompassed themes related to potential downstream complications from testing including false positives and direct complications from invasive procedures, the low likelihood of stress testing altering outcomes, not being worth the cost, and discomfort from stress testing. Individuals who endorsed a low utility construct had significantly lower ITT scores compared to those who did not (2.8 vs. 5.7; P < 0.01).
Strong stimulus (i.e., action cue)
Eight participants cited the strong stimulus of chest pain as being an important factor that influenced their decision-making [Table 3]. Individuals who cited a strong stimulus had higher ITT scores compared to those who did not, but this did not reach statistical significance (6.1 vs. 4.5; P = 0.07).
Relationship between Maximizer-Minimizer Scale and health beliefs
We found no significant associations between MMS and whether participants did or did not endorse a particular theme as something that contributed to their decision-making. The average MMS score of the 15 individuals who said that perceptions of high threat from their condition contributed to their decision-making was not significantly different than the MMS scores of the 45 individuals who did not endorse this theme (4.4 ± 0.7 vs. 4.5 ± 0.9; P = 0.25). Twenty participants said that perceptions of low threat from their condition contributed to their decision-making, but their average MMS score was not significantly different from the 40 participants who did not endorse this theme (4.5 ± 0.9 vs. 4.7 ± 0.8; P = 0.39).
There were 23 participants who said that perceptions of high utility from stress testing contributed to their decision-making, but there was no difference in their MMS score compared to those (n = 37) who did not endorse this theme (4.7 ± 0.9 vs. 4.6 ± 0.8; P = 1.0). Nineteen individuals said that perceptions of low utility from stress testing contributed to their decision-making, but their MMS score was not significantly different from the 41 participants who did not endorse this theme (4.4 ± 1.0 vs. 4.7 ± 0.7; P = 0.20). We did not run the analysis for participants who said that a strong internal stimulus influenced their decision-making because only 8 individuals endorsed it.
Strong minimizers versus strong maximizers
In a post hoc exploratory analysis, we defined strong minimizers by the bottom 10% of MMS scores in the cohort (n = 6), which translated to MMS scores ≤3.5, and strong maximizers by the top 13% of MMS scores (n = 8), which translated to MMS scores >5.5 [Figure 2]. The ITT scores were significantly lower for strong minimizers compared to strong maximizers (3.0 [±1.67] vs. 5.6 [±2.07]; P = 0.02). Five out of 6 (83%) strong minimizers reported being influenced by perceptions of low utility of stress testing compared to only 26% (14/54) of everyone else (P = 0.01) and 3 out of 6 (50%) reported being influenced by both perceptions of low utility and low threat compared to 17% (9/54) of everyone else (P = 0.09). For strong maximizers, 5 of 8 (63%) reported being influenced by perceptions of high utility of stress testing compared to only 35% (18/52) of everyone else (P = 0.24).
| Discussion | | |
In this clinical study of 60 ED patients with low-risk chest pain, a mild correlation was found between MMS-10 score and individuals' intentions to take a cardiac stress test after receiving uniform information about testing, delivered via an iPad tablet. We also found evidence that perceptions related to health beliefs strongly influence decision-making. However, we did not find an association between MMS and health beliefs.
The strength of correlation in our study between MMS and patient preferences is in line with prior studies.[1],[2],[3],[4] One interesting difference in our clinical cohort compared to prior studies is their mean MMS score (4.6 ± 0.8), which was nearly a full point higher than the mean MMS score from prior development and validation studies. Furthermore, our population was composed mainly of maximizers, who made up 75% of the cohort[10],[11],[12] [Figure 2]. The underrepresentation of minimizers begs the question, “Where were the minimizers?” It is unlikely that minimizers experience chest pain less frequently or intensely than maximizers, which suggests that they seek emergent care for the condition less frequently. Perhaps this is good as one certainly does not want to go to the ED more than is necessary and in the case of “low-risk chest pain” it is unlikely that emergent care stemming from the encounter improves outcomes or makes people live longer. However, if minimizers are less likely to go for “low-risk chest pain” then what about for heart attacks? It can be challenging without basic tests like an ECG and bloodwork to distinguish between the two entities. Moreover, if they go less frequently for heart attacks, are they more likely to die suddenly or develop post-MI complications like heart failure that could be prevented with emergent care? If we extend this thinking to its logical conclusion we might ask, “Do strong minimizers experience worse health outcomes or live shorter lives?” Scherer et al.[12] have reported that MMS predicts people's willingness to pursue appropriate care, both when appropriate care means taking high-benefit actions or avoiding low-benefit actions.[3] The distribution of MMS scores in our cohort suggests that individuals with low MMS scores may avoid taking high-benefit action if it involves seeking acute care.
A novel design feature of this study is its use of convergent mixed methods with data transformation to assess the relationship between HBM constructs and preferences for stress testing. To our knowledge, this is the first study to do so. Prior work on the HBM has sought to assess the relationship between HBM constructs and preferences/decision-making using questionnaires and surveys.[13] For the purpose of this study, we believed that qualitative data would provide a richer perspective than what could be achieved through surveys.
There are several important limitations of this study. First, the sample size is small compared to prior studies of MMS that rely on surveys in nonclinical scenarios. This reflects the inherent challenges of recruiting clinical patients during restricted hours when research staff is working and conducting qualitative research (e.g., interviewing, transcribing, coding, and analyzing). Due to the limited sample size, we did not find any statistically significant associations between MMS and HBM constructs; however, our post hoc exploratory findings suggest that MMS orientation, particularly at the tails of the distribution, influences the formation of perceptions related to both threat and utility. The findings from this study will support future research efforts aimed at testing this particular hypothesis, which we estimate would require a sample size two to four times larger if conducted in a similar patient cohort.
Another important limitation to our design is the use of a questionnaire to assess patients' ITT a cardiac stress test after viewing the informational video as opposed to directly assessing whether they did or did not take a cardiac stress test. To this extent, it could be argued that our methods are not significantly different from prior studies that tested MMS using surveys in nonclinical cohorts. We considered this carefully in our design and decided to use ITT test instead of receipt of testing due to several factors. Our research team was not involved in the provision of direct patient care nor did we place any disposition orders on study participants. Thus, after patients completed procedures for this research study, they still had to finalize their disposition with their primary EM provider(s). Because there is a bias among some EM providers, particularly in the US, to encourage patients to undergo inpatient stress testing, we felt this interaction would override patients' natural preferences, at least in some instances, and interfere with our ability to detect a significant association if it existed. Research has shown that many patients will follow the recommendations of providers, even when it conflicts with their preferences.[16]
Despite this potential limitation, we would push back against criticisms that this study design is no different than a survey study in a nonclinical cohort. Nonclinical cohorts have little to no emotion involved in completion of vignette surveys, and it is recognized that emotions influence decision-making, reduce cognitive fixation, and enhance attention.[17] Thus, participants in our study are far different from those represented in previous work involving MMS and support the notion that even in an emotionally charged setting, like chest pain in the emergency room, MMS retains its association with patient preferences in a directional manner.
Finally, it is important to consider the influence of the information provided in the educational video on patient's intentions to take a stress test. The particular percentages for total positive tests, true-positive and false-positive results were based on the experience of the institution where the study was conducted, which had been previously reported.[18] These will vary from center to center due to particulars of the patient populations, the type of stress test used, and specific testing protocols. While the exact percentages may vary and there remains debate about whether stress testing reduces events like MI in this patient population (we allowed for the possibility that it could be in our video), we do not feel that it mitigates the associations reported here or their potential importance. It is already known that presenting information on stress testing to patients in this setting changes their preferences and will reduce the number of those who take a stress test.[9] Thus, the goal of this research was not to develop or validate a decision support tool. Rather, the main intent of the video was to ensure that all patients received the same information and then to try to understand how particular pieces of information influenced patients' perceptions and desire to undergo testing and how it may be related to MMS.
| Conclusions | | |
MMS-10 scores and health beliefs correlated with ED patient's preferences for taking a cardiac stress test after receiving information about the risk of their condition and the probability of outcomes following stress testing. However, a connection was not established between MMS and health beliefs as would be predicted from typical hypothetical models used to represent the HBM [Figure 1]. This may have been due to small sample size and underrepresentation of minimizers in our cohort, which is a separate issue that deserves further attention. In an exploratory, post hoc analysis, we did find associations between MMS and several health beliefs when strong minimizers and maximizers were compared to the rest of the cohort, but this should be viewed as hypothesis generating only. More research is needed to understand how framing of information in the health-care space may interact with stable personality traits to influence decision-making.
Financial support and sponsorship
This research was supported by a grant from the PA Department of Health Tobacco Settlement Fund.
Conflicts of interest
There are no conflicts of interest.
| References | | |
| 1. | |
| 2. | Foy AJ, Liu G, Davidson WR Jr., Sciamanna C, Leslie DL. Comparative effectiveness of diagnostic testing strategies in emergency department patients with chest pain: An analysis of downstream testing, interventions, and outcomes. JAMA Intern Med 2015;175:428-36.  |
| 3. | Sandhu AT, Heidenreich PA, Bhattacharya J, Bundorf MK. Cardiovascular testing and clinical outcomes in emergency department patients with chest pain. JAMA Intern Med 2017;177:1175-82. |
| 4. | Reinhardt SW, Lin CJ, Novak E, Brown DL. Noninvasive cardiac testing vs. clinical evaluation alone in acute chest pain: A secondary analysis of the ROMICAT-II Randomized Clinical Trial. JAMA Intern Med 2018;178:212-9. |
| 5. | Frisoli TM, Nowak R, Evans KL, Harrison M, Alani M, Varghese S, et al. Henry Ford HEART Score Randomized Trial: Rapid discharge of patients evaluated for possible myocardial infarction. Circ Cardiovasc Qual Outcomes 2017;10:e003617. |
| 6. | Foy AJ, Sciamanna C, Davidson WR. Noninvasive testing in emergency department patients with low-risk chest pain: Does the evidence support current guidelines? Cardiol Rev 2016;24:268-72. |
| 7. | Foy AJ, Dhruva SS, Mandrola J. For the patient with “low-risk chest pain”-how low is low? Acad Radiol 2016;23:1587-91. |
| 8. | Safavi KC, Li SX, Dharmarajan K, Venkatesh AK, Strait KM, Lin H, et al. Hospital variation in the use of noninvasive cardiac imaging and its association with downstream testing, interventions, and outcomes. JAMA Intern Med 2014;174:546-53. |
| 9. | Hess EP, Knoedler MA, Shah ND, Kline JA, Breslin M, Branda ME, et al. The chest pain choice decision aid: A randomized trial. Circ Cardiovasc Qual Outcomes 2012;5:251-9. |
| 10. | Scherer LD, Caverly TJ, Burke J, Zikmund-Fisher BJ, Kullgren JT, Steinley D, et al. Development of the medical maximizer-minimizer scale. Health Psychol 2016;35:1276-87. |
| 11. | Scherer LD, Kullgren JT, Caverly T, Scherer AM, Shaffer VA, Fagerlin A, et al. Medical maximizing-minimizing preferences predict responses to information about prostate-specific antigen screening. Med Decis Making 2018;38:708-18. |
| 12. | Scherer LD, Shaffer VA, Caverly T, DeWitt J, Zikmund-Fisher BJ. Medical maximizing-minimizing predicts patient preferences for high- and low-benefit care. Med Decis Making 2020;40:72-80. |
| 13. | Janz NK, Becker MH. The health belief model: A decade later. Health Educ Q 1984;11:1-47. |
| 14. | Amsterdam EA, Wenger NK, Brindis RG, Casey DE Jr., Ganiats TG, Holmes DR Jr., et al. 2014 AHA/ACC guideline for the management of patients with non-ST-elevation acute coronary syndromes: Executive summary: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation 2014;130:2354-94. |
| 15. | Glaser BG. The constant comparative method of qualitative analysis. Soc Probl 1965;12:436-45. |
| 16. | Mendel R, Traut-Mattausch E, Frey D, Bühner M, Berthele A, Kissling W, et al. Do physicians' recommendations pull patients away from their preferred treatment options? Health Expect 2012;15:23-31. |
| 17. | Kozlowski D, Hutchinson M, Hurley J, Rowley J, Sutherland J. The role of emotion in clinical decision making: An integrative literature review. BMC Med Educ 2017;17:255. |
| 18. | Foy A, Baquero GA, Naccarelli GV, Kozak M. Applying the diamond criteria could improve utilization of stress echocardiography for patients who present to the emergency department with low-risk chest pain. Crit Pathw Cardiol 2014;13:49-54. |
|
Search Pubmed for