Do Physicians Prefer Natural Drugs?
The Natural Versus Synthetic Drug Bias in Physicians
Abstract
Abstract.Background: There is a bias for natural versus synthetic drugs in general populations. Aims: We investigated whether physicians who have advanced medical and scientific training and routinely prescribe drugs exhibit this bias. Methods: Physicians and non-physicians were presented with a hypothetical medical situation in which pharmacological therapy was required. Participants were asked if they would prefer a natural or synthetic drug for treatment. Physicians were also asked which drug they would prescribe to a patient. Results: In a forced-choice paradigm, non-physicians (87.5%) and physicians (79.2%) had an equally strong bias for the natural drug, with physicians (74.3%) also preferring the natural drug for patients. When a 9-point drug choice scale was used, including a “no preference” choice (5), non-physicians (M = 6.91) and physicians (M = 5.41) again showed a preference for the natural drug compared to the mid-point of the scale, but the non-physicians’ bias was stronger. Physicians no longer preferred the natural drug for patients (M = 5.15). Limitations: The participants do not represent a random sample and therefore may not represent physicians/non-physicians in general. Additionally, the responses were hypothetical and may not represent behavior in actual medical contexts. Conclusion: These data indicate that physicians and non-physicians exhibit a bias for natural drugs, with physicians also demonstrating a bias for prescribing natural drugs. However, the bias is reduced in physicians compared to non-physicians when a “no preference” option is available, suggesting that advanced medical and/or scientific training may be beneficial in minimizing this bias.
The terms “natural” and “nature” are used frequently in product and service names. For example, natural vitamins, natural hot dogs, natural lawn care services, natural candy, etc. Such terms may reflect the ingredients or processes involved in creating and/or delivering the products or services, but they are also likely to capitalize on the perception that natural products and services are better than unnatural, synthetic, or artificial products and services. For example, Rozin and colleagues (2004) showed that people preferred foods described as natural compared to those described as being processed or human-made. Multiple studies have demonstrated that this naturalness bias exists when considering a diversity of items, including cigarettes (Czoli & Hammond, 2014), meat (Siegrist et al., 2018), personal care products (Apaolaza et al., 2014), and soda (Skubisz 2017), and such preferences hold even when the natural and non-natural items are described as being identical. Meier, Dillard, et al. (2019) suggest that this naturalness bias may be driven by a natural-is-better default belief as well as the belief that natural items are safer than non-natural items. Although natural products may be beneficial, it is inaccurate to assume that anything labeled “natural” is inherently safer or more effective than a synthetic counterpart (Scott & Rozin, 2020). For example, botulinum toxins and arsenic are natural but highly toxic, underscoring the importance of considering the safety and efficacy profiles of items independently from their origin.
Researchers have previously examined the naturalness bias in the context of behavioral medicine and decision making (Meier & Lappas, 2016; Meier, Osorio, et al., 2019). In a series of studies (Meier & Lappas, 2016), preferences for natural versus synthetic drugs in a medical context were examined, and it was demonstrated that people preferred a natural to a synthetic drug for both minor and serious hypothetical medical conditions even though the drugs were described as equally safe and effective. Notably, it was also found that a significant percentage of people preferred a natural drug even when it was described as less safe or less effective than a synthetic drug. Participants in these studies rated natural drugs as safer than synthetic drugs, which seemingly contributed to the natural drug bias. Importantly, it has been shown that the preference for a natural drug can be replicated in a behavioral choice study but that the natural drug bias can be reduced when individuals are educated about the inherent bias for natural drugs (Meier, Osorio, et al., 2019). Much of the work investigating the natural drug bias has been conducted in Western cultures, but studies have shown that the bias exists in Eastern cultures as well and can be predicted by individual differences in connectedness to nature or religiosity (Cao & Li, in press; Li & Cao, 2020).
A natural drug bias is significant within the context of health behavior because this bias may influence the pharmacological choices made by patients and possibly their physicians, with potentially significant health ramifications. For example, individuals may erroneously assume that a drug labeled “natural” is inherently safe and therefore may not be vigilant about monitoring for potential adverse side effects. Conversely, individuals may be hesitant to utilize a beneficial drug or treatment if it is not described as natural. For example, compared to individuals with a less pronounced bias, individuals with a stronger natural versus synthetic drug bias reported that they would be less likely to take a COVID-19 vaccine (Meier et al., in press). Additionally, individuals who reported preferring a natural versus synthetic treatment for influenza were less likely to get the influenza vaccine (DiBonaventura & Chapman, 2008).
Many factors influence an individual’s decision to initiate and/or adhere to a pharmacological regimen. One important factor in the medical decision-making process is the interaction between patients and physicians. Physicians give advice, make recommendations, and prescribe drugs. It is of interest, therefore, to examine if physicians exhibit a natural drug bias, as such a bias could have significant consequences for the medical advice and care received by patients. Although physicians, by means of their training and experience, are well versed in both basic science and pharmacological science, the naturalness bias is strong across a host of participant samples and items (Meier, Dillard, et al., 2019). Furthermore, physicians have been shown, as people in general, to exhibit several cognitive biases (e.g., overconfidence, the anchoring effect, etc.; Saposnik et al., 2016). Therefore, it is reasonable to hypothesize that the natural drug preference may be evident in physicians, although possibly tempered by their scientific and medical knowledge. Indeed, a prior study showed that in a sample of obstetricians and gynecologists, approximately 31% preferred to use a natural hormone replacement therapy compared to a human-made therapy even when the treatments were described as identical (3% preferred the human-made, and 66% had no preference), suggesting that such a bias may exist (Baron et al., 1998). However, a systematic investigation of the natural drug bias in physicians is necessary before making conclusions in this domain.
Methods
Overview of Studies, Data/Ethics Statements, and Statistical Power Statement
In two studies (total N = 411), we examined whether physicians demonstrate a preference for natural versus synthetic drugs for themselves and/or for their patients. In Study 1, a forced-choice paradigm was used, and in Study 2, a drug choice rating paradigm was used. Beliefs about safety and efficacy were also examined.
In both studies, we report all measures, conditions, data exclusions, and the manner in which we determined sample sizes. All study materials and the SPSS data files are available at an Open Science Framework webpage: https://osf.io/mxy2q/. Both studies were approved by the Lebanon Valley College Institutional Review Board (IRB PROTOCOL ID#: 2020-49). Informed consent was obtained from all individual participants included in the study.
In five studies examining the natural versus synthetic drug bias, Meier and Lappas (2016) found medium to large effect sizes when examining preferences for natural versus synthetic drugs. The current studies examined both drug choice (natural vs. synthetic) and sample type (physicians vs. non-physicians). Due to the expected difficulty in collecting data from a very large sample of physicians, the goal of this work was to include a minimum of 100 physicians in each of Studies 1 and 2, and to compare the sample of physicians in each study to at least 100 non-physicians for a total of 200 participants in each study (400 total participants).
In terms of replicating the natural drug bias in each sample in each study, the sample sizes we sought to collect (100 per sample type) gave us 80% power to detect a medium effect in Study 1 (Cramer’s ϕ = .28) and a small to medium effect in Study 2 (d = .28). In terms of examining differences between groups in each study, the sample sizes gave us 80% power in Studies 1 and 2 to detect a small to medium effect (Study 1: Cramer’s ϕ = .20; Study 2: d = .40).
Study 1
Participants
Physicians
The US News and World Report ranking of the top US medical schools in primary care were used to identify schools to collect email addresses for physicians in internal medicine, primary care, general medicine, etc. We identified email addresses by going to each school’s directory of physicians and then determining if an email address was provided for physicians. Most schools included email addresses, but some did not. Physicians were contacted via email and asked to consider completing a 3-minute survey about their opinions on drugs. The number of physicians contacted at any single school was limited to approximately 100; when the email addresses of more than 100 physicians were available, individuals were randomly selected. Follow-up emails were sent 7 days later. We anticipated a low response rate for several reasons, including the busy schedule of physicians (especially during a pandemic), the fact that our email was coming from a source unknown to the physicians and might therefore be perceived as spam, and because we did not offer an incentive for completing the study. A total of 1,526 physicians were emailed in Study 1, and 105 questionnaire responses were received, which was a rate better than expected given the issues mentioned above. Across both studies, we emailed physicians from 30 different schools in 24 different US states.
The 105 physicians (68 females, 34 males, 1 transgender, 2 non-report) had a mean age of 52.61 (SD = 11.72) years. The majority of physicians self-identified as White (79), and the remaining physicians self-identified as Asian (11), Latinx (4), Black (3), Multiracial (3), or Hispanic (1). Four physicians did not report race.
One physician was removed for taking an excessive amount of time on the questionnaire (more than three SDs from the mean questionnaire completion time) and three physicians were removed for not completing multiple questions related to the scenarios. The final sample size was 101.
Non-Physicians
One hundred ten non-physicians were recruited from https://www.prolific.co/ within the same dates of the physicians’ study. Prolific.co is a crowdsourcing website with tens of thousands of participants used in marketing and behavioral research. We selected participants who were located in the US, were of US nationality, and spoke English as a first language. Participants were paid $0.45 for the 3-minute study. The 110 non-physicians (46 females, 60 males, 3 non-binary, 1 non-report) had a mean age of 31.47 (SD = 12.11) years. The majority of non-physicians self-identified as White (77), and the remaining non-physicians self-identified as Asian (10), Black (8), Latinx (5), Multiracial (5), Hispanic (4), or Native Hawaiian (1). Non-physicians’ reported education level was as follows: high school graduate (18), some college (30), 2-year college or technical degree (11), 4-year degree (37), some postgraduate work (6), and postgraduate degree (8).
Six non-physicians were removed because they reported having an MD degree. The final sample size was 104.
Materials and Procedure
After giving informed consent, participants read a scenario that was based upon past work by Meier and colleagues (Meier & Lappas, 2016; Meier, Osorio, et al., 2019):
Imagine that you learn that you have a medical condition and you need to take a drug to treat it. You have to choose between one of the two options shown below:
Option 1 is a synthetic drug made from ingredients NOT FOUND in nature. Studies have been conducted on this drug for 20 years. It has been shown to be effective in 85% of users. The drug has also been shown to cause mild side effects on rare occasions and serious side effects in 0.5% of users.
Option 2 is a natural drug made from ingredients FOUND in nature. Studies have been conducted on this drug for 20 years. It has been shown to be effective in 85% of users. The drug has also been shown to cause mild side effects on rare occasions and serious side effects in 0.5% of users.
Which drug would you take?
After participants chose one drug, they rated how safe (1 = not at all safe to 9 = very safe) and effective (1 = not at all effective to 9 = very effective) they perceived each drug to be using a 9-point scale. Next, physicians were asked the same question above about the natural versus synthetic drug but this time in the context of treating a patient as a caregiver. They were asked which drug they would prescribe. All participants answered demographic questions including age, gender, race, highest level of educational attainment (some high school, high school graduate, some college, 2-year or technical degree, 4-year degree, some graduate work, postgraduate degree), and whether or not they were medical doctors. Finally, participants were debriefed about the true nature of the study.
Study 2
Participants
Physicians
Emails were sent to physicians using the same process described in Study 1. A total of 936 physicians were emailed in Study 2, and 108 questionnaire responses were received. As mentioned above, across both studies, we emailed physicians from 30 different schools in 24 different US states.
The 108 physicians (41 females, 59 males, 8 non-report) had a mean age of 50.61 (SD = 10.71) years. The majority of physicians self-identified as White (83), and the remaining physicians self-identified as Asian (10), Black (2), Latinx (1), or Multiracial (1). Eleven physicians did not report race.
Seven physicians were removed for not completing multiple questions related to the scenarios, and one physician was removed for taking an excessive amount of time on the questionnaire (more than three SDs from the mean questionnaire completion time). The final sample size was 100.
Non-Physicians
One hundred eleven non-physicians were recruited from https://www.prolific.co/ within the same dates of the physicians’ study. We selected participants who were located in the US, were of US nationality, and spoke English as a first language. Participants were paid $0.45 for the 3-minute study. The 111 non-physicians (48 females, 58 males, 3 non-binary, 2 transgender) had a mean age of 31.83 (SD = 11.74) years. The majority of non-physicians self-identified as White (81), and the remaining non-physicians self-identified as Asian (13), Black (6), Latinx (6), Multiracial (4), or Native American (1). Non-physicians’ reported education level was as follows: some high school (2), high school graduate (14), some college (30), 2-year college or technical degree (10), 4-year degree (35), some postgraduate work (3), and postgraduate degree (17).
Four non-physicians were removed because they reported having an MD degree, and one non-physician was removed for taking an excessive amount of time on the questionnaire (more than three SDs from the mean questionnaire completion time). The final sample size was 106.
Materials and Procedure
After giving informed consent, participants completed the same drug choice question used in Study 1, but the response mode was different. Participants answered using a drug choice rating scale: 1 = I strongly prefer the synthetic drug; 2; 3 = I moderately prefer the synthetic drug; 4; 5 = I have no preference between the two drugs; 6; 7 = I moderately prefer the natural drug; 8; 9 = I strongly prefer the natural drug.
After participants made a rating choice, they rated the safety and effectiveness of each drug as in Study 1. Next, Physicians were asked the same question above about the natural versus synthetic drug, but this time in the context of treating a patient as a caregiver; they were asked which drug they would prescribe, and they used the 9-point scale shown above. Finally, all participants answered the same demographic questions used in Study 1 and were then debriefed about the true nature of the study.
Results
Study 1
An unbiased finding would be one in which each sample of participants chose the drugs at an equal rate (50% for each drug), given that no additional information was provided. We first examined drug choice separately for each sample. As shown on the left side of Figure 1, 91 of 104 (87.5%) non-physicians chose the natural drug, and 13 of 104 (12.5%) chose the synthetic drug. This difference was statistically different from a 50:50% split and illustrates a strong natural drug bias, χ2(1, N = 104) = 58.50, p < .001, Cramer’s ϕ = .75. As shown on the right side of Figure 1, 80 of 101 (79.2%) physicians chose the natural drug, and 21 of 101 (20.8%) chose the synthetic drug. This difference was statistically different from a 50:50% split and again illustrates a strong natural drug bias, χ2(1, N = 101) = 34.47, p < .001, Cramer’s ϕ = .58. The difference in drug choice between physicians and non-physicians was not significant, χ2(1, N = 205) = 2.55, p = .11, Cramer’s ϕ = .111.
Physicians answered the same hypothetical question regarding drug choice in the context of prescribing a drug as a care provider to a patient. Seventy-five of 101 (74.3%) physicians chose the natural drug, and 26 of 101 (25.7%) chose the synthetic drug. This difference was statistically different from a 50%:50% split and again illustrates a strong natural drug bias, χ2(1, N = 101) = 23.77, p < .001, Cramer’s ϕ = .49.
We next examined safety and effectiveness ratings. The mean and standard deviations by sample type are shown in Table 1.
As in prior work (Meier & Lappas, 2016; Meier, Osorio, et al., 2019), we created two different scores by subtracting synthetic drug ratings from natural drug ratings for both safety and effectiveness. Positive numbers indicate that participants perceived the natural drug to be safer or more effective than the synthetic drug, whereas negative numbers indicate that participants perceived the synthetic drug to be safer or more effective than the natural drug. Neither of these difference scores was normally distributed (Shapiro–Wilk tests all ps < .001). Therefore, we also performed bootstrapping with 5,000 samples in addition to the standard tests reported below.
We used one-sample t-tests to determine if the means were different from 0 for both physicians and non-physicians in terms of the safety and effectiveness difference scores (note: one Physician did not complete these ratings). This analysis revealed that non-physicians rated the natural drug as significantly safer than the synthetic drug (M = 0.76, SD = 1.66), t(103) = 4.66, p < .001, d = .46 (95% CI of the mean difference with bootstrapping: .46, 1.09). The effectiveness difference score was not significantly different from 0 in non-physicians (M = 0.14, SD = 0.99), t(103) = 1.49, p = .140, d = .14 (95% CI of the mean difference with bootstrapping: −.05, .34). Physicians, however, did not rate the natural drug as significantly safer than the synthetic drug (M = 0.00, SD = 0.70), t(99) = 0.00, p = 1.00, d = .00 (95% CI of the mean difference with bootstrapping: −.14, .13). The effectiveness difference score was also not significantly different from 0 in physicians (M = −0.08, SD = 0.49), t(99) = −1.65, p = .103, d = −.17 (95% CI of the mean difference with bootstrapping: −.18, .00).
Finally, we compared the difference scores for both safety and effectiveness ratings between physicians and non-physicians using independent-samples t-tests. The safety rating difference score was significantly higher in non-physicians (M = 0.76, SD = 1.66) than physicians (M = 0.00, SD = 0.70), t(202) = 4.22, p < .001, d = .59 (95% CI of the mean difference with bootstrapping: .41, 1.10). The effectiveness rating difference score was also significantly higher in non-physicians (M = 0.14, SD = 0.99) compared to physicians (M = −0.08, SD = 0.49), t(202) = 2.04, p = .042, d = .29 (95% CI of the mean difference with bootstrapping: .01, .44).
The results of Study 1 revealed that physicians had the same preference for natural versus synthetic drugs as non-physicians even though the drugs were said to be equally safe and effective. Physicians also showed this bias when thinking about prescribing a natural versus synthetic drug for a hypothetical patient. However, physicians did not rate the natural drug as safer than synthetic drugs even though non-physicians did. Non-physicians overlooked the fact that natural and synthetic drugs were said to be equally safe, but physicians did not. This finding may suggest that physicians are more analytical when evaluating drug safety data. Yet, physicians still showed a preference for the natural drug when forced to choose one.
Study 2
In Study 2, we sought to replicate the findings from Study 1 with one major difference. In Study 1, a forced-choice paradigm required participants to choose either a natural or synthetic drug. This paradigm is useful because it mimics situations encountered in clinical settings in which a choice must be made, as is the case when deciding upon a drug. The forced-choice paradigm did not allow participants to state that either drug choice was acceptable because in clinical settings, even though any one of multiple drugs might be appropriate for treatment, ultimately, the patient and/or physician must make a selection, and this choice often depends upon individual preferences (Elwyn et al., 2009). To further investigate the extent of the natural drug bias, however, in Study 2, we altered our paradigm to allow participants to choose a “no preference” option when prompted to select a natural or synthetic drug.
Within the paradigm of Study 2, an unbiased finding would be one in which each sample of participants had a mean rating that was not different than 5 (“I have no preference between the two drugs”). As in Study 1, the continuous variables did not follow a normal distribution (Shapiro–Wilk tests all ps < .001). We, therefore, performed bootstrapping with 5,000 samples in addition to the standard tests reported below. We first examined drug choice ratings separately for each sample. The non-physicians’ mean drug choice rating (M = 6.91, SD = 1.71) was significantly higher than the scale mid-point of 5, t(105) = 11.44, p < .001, d = 1.11 (95% CI of the mean difference with bootstrapping: 1.57, 2.22). Similarly, the physicians’ mean drug choice rating (M = 5.41, SD = 1.07) was significantly higher than the scale mid-point of 5, t(99) = 3.82, p < .001, d = .38 (95% CI of the mean difference with bootstrapping: .20, .62). Unlike Study 1, as shown in Figure 2, the difference in the mean drug choice ratings between physicians and non-physicians was significant, t(204) = 7.45, p < .001, d = 1.04 (95% CI of the mean difference with bootstrapping: 1.11, 1.89). Non-physicians reported a much stronger preference for natural versus synthetic drugs compared to physicians.
Physicians’ responses to the question regarding the drug choice rating for a hypothetical patient was not significantly different from 5, (M = 5.15, SD = 1.00), t(99) = 1.50, p = .136, d = .15 (95% CI of the mean difference with bootstrapping: −.05, .35).
As in Study 1, we examined safety and effectiveness ratings. The mean and standard deviations by sample type are shown in Table 2.
We again calculated difference scores for the safety and effectiveness ratings. We used one-sample t-tests to determine if these scores were different from 0 for both physicians and non-physicians. This analysis revealed that non-physicians rated the natural drug as significantly safer than the synthetic drug (M = 0.69, SD = 1.56), t(105) = 4.53, p < .001, d = .44 (95% CI of the mean difference with bootstrapping: .41, 1.00). The effectiveness difference score was not significantly different from 0 in non-physicians (M = 0.06, SD = 1.14), t(105) = 0.51, p = .609, d = .05 (95% CI of the mean difference with bootstrapping: −.14, .29). Physicians did not rate the natural drug as significantly safer than the synthetic drug (M = −0.04, SD = 0.65), t(99) = −0.62, p = .540, d = −.06 (95% CI of the mean difference with bootstrapping: −.18, .08). The effectiveness difference score was also not significantly different from 0 in physicians (M = −0.02, SD = 0.20), t(99) = −1.00, p = .320, d = −.10 (95% CI of the mean difference with bootstrapping: −.08, −.02).
Finally, we compared the difference scores for both safety and effectiveness ratings between physicians and non-physicians. We used independent-samples t-tests. As found in Study 1, the safety rating difference score was significantly higher in non-physicians (M = 0.69, SD = 1.56) compared to physicians (M = −0.04, SD = 0.65), t(204) = 4.32, p < .001, d = .60 (95% CI of the mean difference with bootstrapping: .42, 1.07). The effectiveness rating difference score was not significantly different in non-physicians (M = 0.06, SD = 1.14) compared to physicians (M = −0.02, SD = 0.20), t(204) = 0.66, p = .507, d = .09 (95% CI of the mean difference with bootstrapping: −.13, .30).
The results of Study 2 are somewhat similar to what was found in Study 1. Both physicians and non-physicians preferred the natural versus synthetic drug even when a no preference rating was included in the response options. However, unlike in Study 1 and Study 2, non-physicians showed a much stronger natural versus synthetic drug bias, suggesting that physicians were more likely to consider and select the no-preference option. Non-physicians again rated the natural drug as safer than the synthetic drug, but physicians did not show a safety-rating difference between the two drugs.
Discussion
The results of two studies revealed that both physicians and non-physicians exhibit a bias for natural versus synthetic drugs in hypothetical medical scenarios. This bias occurs even when the drugs are described as identical in terms of safety and effectiveness. The strength of the naturalness bias observed was similar in physicians and non-physicians when a forced-choice option was required (Study 1), but it was significantly greater in non-physicians when a “no preference” option was allowed (Study 2). Physicians still exhibited a natural drug bias overall in Study 2, but it was much weaker than non-physicians. In both studies, non-physicians perceived the natural drug as safer than the synthetic drug, but physicians did not. These results suggest that non-physicians may prefer natural drugs because they perceive them to be safer (Meier & Lappas, 2016; Meier, Osorio, et al., 2019), whereas physicians apparently prefer them for alternate reasons (e.g., the inherently perceived positivity of natural items).
The current studies extend previous work by demonstrating that physicians exhibit the natural drug bias. Not only do physicians possess a bias for natural drugs when considering a hypothetical pharmacological treatment for themselves, but this bias extends to their prescription preferences for a patient. Notably, however, this bias is diminished when a “no-preference” option is available. It is likely that the advanced medical and scientific training and experience of physicians diminishes the bias by enabling them to better analyze the safety and efficacy of the drugs independently from the consideration of the drugs’ origins. Indeed, physicians, unlike non-physicians, did not rate the natural drug as significantly safer than the synthetic option. These findings align with results from previous work showing that when individuals are educated about the naturalness bias and informed that neither natural nor synthetic substances are inherently good or bad, the preference for a natural drug is reduced (Meier, Osorio, et al., 2019). Thus, although a “no preference” option is often not possible in a clinical setting, these results are instructive in demonstrating that education, training, and/or experience are instrumental in tempering the natural drug bias. These findings may be useful when applied to clinical situations in which patients and physicians interact to develop plans for pharmacological interventions. Because the general public often does not possess the same level of medical or scientific knowledge as physicians, it is vital that physicians be conduits of knowledge to their patients by educating them about the natural drug bias and the importance of recognizing that the biological activity of any drug is dependent upon its structure and not its origin. Physicians are often on the front lines of medical decision-making, and our results indicate that although they exhibit a bias for natural drugs, they are better able to defy this bias in certain situations.
Overall, the natural drug bias appears to be one of the several cognitive biases physicians share with non-physicians (Saposnik et al., 2016). Both physicians and non-physicians show a preference for natural drugs, and this preference may have implications in the context of health behavior in terms of drug choice and use. For example, when a natural drug is available as a medical treatment option, both physicians and their patients may prefer that option and be more likely to follow the treatment to its completion. This may have either beneficial or detrimental ramifications. If a drug obtained from natural sources is equally or more safe and effective than a synthetic alternative, the bias for the drug will serve to benefit the patient, and it would be advantageous for the physician to highlight the natural origin of the compound. If, however, a synthetic alternative has a more favorable safety and/or efficacy profile, the patient and physician bias for a natural drug may be detrimental as it could result in an unwise decision to bypass the synthetic drug and prescribe and/or take a less safe and/or effective natural option. The current findings also suggest that if safe and effective natural drug alternatives to currently utilized synthetic compounds were more readily available, patients might be more inclined to commence and adhere to beneficial pharmacological treatment regimens. Short of this, it is important that physicians and patients are aware of their inherent preference for natural drugs and are educated in a manner that allows them to challenge this inherent preference when possible so that the bias does not interfere with prudent medical decision-making for themselves or others.
Limitations
The current results are not without limitations. First, as with similar research of the present type, the physicians and non-physicians included in these studies do not represent random samples across the US. The physicians were located at top medical schools in the US, and the non-physicians were from Prolific.co, which has a large sample of individuals, but both sample types are convenience samples, and therefore responses may not represent physicians and non-physicians in general. Second, the responses made in the current studies were hypothetical in nature and may not represent true behavior. The scenarios illustrate the thought- and decision-making processes of physicians and non-physicians, but the decisions were based upon self-report. Past work has demonstrated that the natural drug bias influences behavioral choices (Meier, Osorio, et al., 2019), but further investigation is needed to ascertain if this would occur with a sample of physicians. Finally, it must be acknowledged that there are many factors in addition to the natural or synthetic label of a drug that may influence a patient and/or physician – for example, drug cost and availability.
These limitations notwithstanding, the results of these studies are informative in revealing a natural drug bias in physicians, and future work aimed at further characterizing this bias and its influence on behavior is warranted.
References
(2014). Natural ingredients claim’s halo effect on hedonic sensory experiences of perfumes. Food Quality and Preferences, 36, 81–86. https://doi.org/10.1016/j.foodqual.2014.03.004
(1998). Attitudes of obstetricians and gynecologists toward hormone replacement therapy. Medical Decision Making, 18(4), 406–411. https://doi.org/10.1177/0272989X9801800408
(in press). Harmony between humanity and nature: Natural vs. synthetic drug preference in Chinese atheists and Taoists. Journal of Religion and Health. https://doi.org/10.1007/s10943-021-01314-6
(2014) Cigarette packaging: Youth perceptions of “natural” cigarettes, filter references, and contraband tobacco. Journal of Adolescent Health, 54(1), 33–39. https://doi.org/10.1016/j.jadohealth.2013.07.016
(2008). Do decision biases predict bad decisions? Omission bias, naturalness bias, and influenza vaccination. Medical Decision Making, 28(4), 532–539. https://doi.org/10.1177/0272989X07312723
(2009). Dual equipoise shared decision making: Definitions for decision and behaviour support interventions. Implementation Science, 4(1), Article
75 . https://doi.org/10.1186/1748-5908-4-75(2020). For the love of nature: People who prefer natural versus synthetic drugs are higher in nature connectedness. Journal of Environmental Psychology, 71, 1–6. https://doi.org/10.1016/j.jenvp.2020.101496
(2021). Study materials and data files for “Do physicians prefer natural drugs? The natural versus synthetic drug bias in physics.” https://osf.io/mxy2q/
(2019). Naturally better? A review of the natural-is-better bias. Social and Personality Psychology Compass, 13, Article
e12494 . https://doi.org/10.1111/spc3.12494(in press). Predictors of the intention to receive a SARS-CoV-2 vaccine. Journal of Public Health. https://doi.org/10.1093/pubmed/fdab013
(2016). The Influence of safety, efficacy, and medical condition severity on natural versus synthetic drug preference. Medical Decision Making, 36(8), 1011–1019. https://doi.org/10.1177/0272989X15621877
(2019). A behavioral confirmation and reduction of the natural versus synthetic drug bias. Medical Decision Making, 39(4), 359–369. https://doi.org/10.1177/0272989X19838527
(2004). Preference for natural: instrumental and ideational/moral motivations, and the contrast between foods and medicines. Appetite, 43(2), 147–154. https://doi.org/10.1016/j.appet.2004.03.005
(2016). Cognitive biases associated with medical decisions: A systematic review. BMC Medical Informatics and Decision Making, 16, Article
138 . https://doi.org/10.1186/s12911-016-0377-1(2020). Actually, natural is neutral. Nature Human Behavior, 4, 989–990. https://doi.org/10.1038/s41562-020-0891-0
(2018). Perceived naturalness and evoked disgust influence acceptance of cultured meat. Meat Science, 139, 213–219. https://doi.org/10.1016/j.meatsci.2018.02.007
(2017). Naturally good: Front-of-package claims as message cues. Appetite, 108, 506–511. https://doi.org/10.1016/j.appet.2016.10.030
