Skip to main content
Download PDF

Decision threshold models in medical decision making: a scoping literature review

BMC Medical Informatics and Decision Making volume 24, Article number: 273 (2024) Cite this article

Abstract

Background

Decision thresholds play important role in medical decision-making. Individual decision-making differences may be attributable to differences in subjective judgments or cognitive processes that are captured through the decision thresholds. This systematic scoping review sought to characterize the literature on non-expected utility decision thresholds in medical decision-making by identifying commonly used theoretical paradigms and contextual and subjective factors that inform decision thresholds.

Methods

A structured search designed around three concepts—individual decision-maker, decision threshold, and medical decision—was conducted in MEDLINE (Ovid) and Scopus databases from inception to July 2023. ProQuest (Dissertations and Theses) database was searched to August 2023. The protocol, developed a priori, was registered on Open Science Framework and PRISMA-ScR guidelines were followed for reporting on this study. Titles and abstracts of 1,618 articles and the full texts for the 228 included articles were reviewed by two independent reviewers. 95 articles were included in the analysis. A single reviewer used a pilot-tested data collection tool to extract study and author characteristics, article type, objectives, theoretical paradigm, contextual or subjective factors, decision-maker, and type of medical decision.

Results

Of the 95 included articles, 68 identified a theoretical paradigm in their approach to decision thresholds. The most common paradigms included regret theory, hybrid theory, and dual processing theory. Contextual and subjective factors that influence decision thresholds were identified in 44 articles.

Conclusions

Our scoping review is the first to systematically characterizes the available literature on decision thresholds within medical decision-making. This study offers an important characterization of the literature through the identification of the theoretical paradigms for non-expected utility decision thresholds. Moreover, this study provides insight into the various contextual and subjective factors that have been documented within the literature to influence decision thresholds, as well as these factors juxtapose theoretical paradigms.

Peer Review reports

Background

Decision thresholds play an important and understudied role in decision-making. Often, decision thresholds are viewed as the “linchpin” between evidence and decision-making [1, 2]. The concept of a decision threshold is often familiar to those with a basic understanding of the judicial system where standards of proof assume different evidentiary cut-offs such as, “‘beyond a reasonable doubt’ (typical of criminal court) and the less-stringent ‘preponderance of evidence’ (typical of a civil court)” [1]. Decision thresholds, be it for individuals or society, reflect the perceived value of the consequences (e.g., convicting an innocent person to prison versus letting a guilty person walk free) [1, 3,4,5,6,7]; effectively, a decision threshold dichotomizes a decision into taking or not taking an action [8,9,10,11]. Generally, differing decisions are attributed to individual differences in subjective judgements that are attributed to different decision thresholds [12]. Differences in assessments of the benefits and harms of particular decisions, or subjective risk-assessment judgements, help to explain the variation in decision thresholds [13, 14]. Within the context of medical decision-making, Djulbegovic and colleagues assert that the “development of [the] threshold model is considered as one of the most important advances in medical decision-making” [15]. Understanding the ways that individuals incorporate subjective judgements and values into decision thresholds [7] and the cognitive mechanisms they employ to inform decisions could suggest opportunities to understand variation within medical decision-making [8, 13].

While arguably of central interest to the decision making process, surprisingly little research is dedicated to decision thresholds [16]. Within the existing literature, several different theoretical paradigms have been identified as underpinning how an individual might arrive at a decision threshold [9]. A narrative review by Djulbegovic and colleagues [14] summarized theoretical paradigms underpinning decision thresholds that inform medical decision-making: 1) expected utility theory (EUT) models [10, 11], 2) regret based decision models [8, 17,18,19,20,21], 3) dual processing/dual system models [12, 22], and 4) other models [14]. Djulbegovic and colleagues argued/asserted that theoretical paradigms for decision thresholds have not been adequately explored or employed empirically, nor has the concept been meaningfully integrated into theoretical investigations of clinical decision-making [14]. Despite the perceived importance of decision thresholds in medical decision-making, the various papers bearing upon the topic have yet to be systematically reviewed. While EUT threshold models are well established within the medical decision-making literature (see Pauker and Kassirer [10, 11]) [14], a considerable body of literature recognizes that in practice different underpinning paradigms appear to be plausible and, on occasion, better matches to practice [14, 23, 24]. EUT and non-EUT paradigms differ in how they weigh probabilities and scale utilities [2]. Typically, an EUT approach does not account for contextual and subjective factors that inform an individual’s decision-making, whereas non-EUT paradigms often account for affect, emotion, values, preferences and other contextual/ situational factors [2]. We chose to exclude EUT decision threshold paradigms from our systematic scoping review because a considerable body of literature has recognized that people do not always make decisions with a goal of maximizing expected utility [23,24,25,26]. For example, Djulbegovic and colleagues (2014) investigated which decision threshold paradigms (i.e., EUT, regret and dual-processing theory) most accurately reflected physicians’ actions for pulmonary embolisms and acute myeloid leukemia [25]. Results from this study suggested that the EUT paradigm for decision threshold was a weaker predictor of physician decisions than regret-based and dual processing theory paradigms (the latter of which was the best predictor) [25].

We employed a scoping review approach to systematically review alternative theoretical paradigms underpinning decision thresholds. We seek to categorize the non-EUT paradigms (i.e., those do not share the same general structure as described in Pauker and Kassirer) [10, 11]. This scoping review adds to the body of knowledge by summarizing the most recent research on decision thresholds within medical decision-making, and by extending this corpus of knowledge to include literature on the paradigms that physicians and non-physicians (e.g., allied health professionals, lay-persons/ patients, etc.) use for non-EUT decision thresholds in medical decision-making. The following two objectives were driving the review described in this paper:

  • Objective 1: Identify and categorize what theoretical paradigms have been developed to establish or estimate decision threshold(s) within medical decision-making.

  • Objective 2: Identify what consideration an individual’s subjective judgment(s) (e.g., attitudes, emotions, preferences, risk perceptions, values) have been given in a context of decision threshold(s) within medical decision-making.

Methods

Following Johanna Briggs Institute (JBI) methodology [27], we conducted a scoping review to systematically identify the non-EUT theoretical paradigms underpinning medical decision thresholds. The protocol, developed a priori, was registered on Open Science Framework [28] and PRISMA-ScR guidelines [29] were followed for reporting on this study. Our methods are summarized here. Full methodological details are included in the supplementary text.

Data sources

We designed a search strategy around three concepts: individual decision-maker, decision threshold, and medical decision. We sought to identify articles where the term “decision threshold” and other like terms appeared within the title or abstract of an article in the context of “medical decisions” or other like terms. The search was piloted in MEDLINE (Ovid) and subsequently translated for Scopus search syntax. Both searches were completed in July, 2023. The ProQuest (Dissertations and Theses) database was searched in August 2023 using the phrase “decision threshold”. Results of the search were deduplicated using the Zotero Duplicate Merger Plug-In (https://www.zotero.org/). We used Covidence software (https://www.covidence.org/) to facilitate article review and selection.

Study selection

We selected articles published in English that acknowledged theoretical paradigms for decision thresholds, and where medical or health related decisions were made by individuals (e.g., physicians, allied health professionals, lay individuals, etc.). We excluded articles that applied decision threshold models at the policy level (e.g., cost-effectiveness analysis and willingness to pay thresholds).

Since we wished to categorize the theories that have been conceptualized as underpinning decision thresholds (objective 1), our screening criteria were deliberately broad. As the literature on decision thresholds transcends disciplinary boundaries, we kept the title and abstract screening criteria purposefully conceptual rather than concrete to allow for fields using different terms to capture the same or similar ideas. We piloted screening criteria, summarized in Table 1, in a subset of studies prior to implementation and achieved high interrater agreement. Titles and abstracts of articles identified in our search were screened by two independent reviewers (AS and AC) and disagreements were discussed and resolved through consensus. Full text of articles passing initial screening were reviewed by the two independent reviewers (AS and AC) using the full text screening criteria summarized in Table 1 following a pilot screening process where, again, high inter-rater agreement was achieved. Disagreements on full text screening assessments were resolved through a consensus discussion.

Table 1 Summary of Inclusion/ Exclusion Criteria Based on Screening Phase
Full size table

Reference lists of all articles included after full-text review were screened using title and abstract criteria. Potentially suitable articles were assessed using full-text inclusion and exclusion criteria.

A total of 2,358 articles were identified through database and reference list searching. After de-duplication, 1,618 articles entered the two-stage screening process. Ultimately, 95 articles were included in the review. Details of study selection are depicted in Fig. 1.

Fig. 1
figure 1

PRISMA flow diagram

Full size image

Data extraction

A data extraction form designed to capture details about each article was pilot tested by a single researcher (AS) on a subset of included articles. Team discussions helped to streamline the form. Extracted characteristics included: author details, title of the journal, country of study (where applicable), the objective of the article, article type (e.g., empirical, conceptual/ theoretical, review, etc.), theoretical foundations (if any), contextual/ subjective factors (if any), types of decision threshold (e.g., treat/ not treat; test/ not test, etc.), decision threshold mathematical formulation (if any), the decision-maker (e.g., physician, nurse, patient, guardian/ caregiver, etc.), type of medical decision (e.g., coronary artery disease, stroke, COPD, etc.), authors’ conclusions, as well as relevant references (see Identifying Additional Literature). All data extraction was performed by a single researcher (AS); team discussions were called for difficult or ambiguous extractions.

Identifying additional literature

The reference lists of all articles that met the full-text inclusion criteria were scanned using title and abstract criteria and potentially suitable articles were assessed using full-text inclusion/exclusion criteria. Data was charted from included articles as detailed above.

Synthesis of results

Selected characteristics of the included articles are tabulated in the supplementary text. Extracted data were evaluated thematically to identify the theoretical paradigms represented in the literature. We narratively summarized the literature within each paradigm and illustrate the distribution of records within each paradigm in tabular form. We also tabulated summary data from articles that discussed contextual and subjective factors associated with decision thresholds and summarize these narratively.

Results

Of the 95 included articles, 27 discussed the concept of decision thresholds in medical decision-making applications but did not reference or explicitly state a theoretical paradigm as conceptually underpinning the decision threshold, while the remaining 68 did. Here we discuss these remaining 68 articles (Table 2).

Table 2 Common Theoretical Paradigms of Non-Expected Utility Decision Thresholds
Full size table

The most common ex-ante theoretical paradigms (i.e., estimating a decision threshold prior to the decision being made) used to determine an individual’s decision threshold are regret theory, hybrid theory, and dual processing theory. The most common ex-post methodological approach (i.e., estimating a decision threshold after a decision has been made) used to estimate an individual’s decision threshold involves regression techniques (e.g., ordinal, logistical, etc.).

Regret theory

Regret theory is the most common theoretical paradigm for non-EUT decision thresholds (mentioned/discussed in 24 out of 68 articles~ 35%). Shortly after regret theory was first propounded [23, 24, 89], Feinstein [35] tailored the concept to the medical decision-making setting by identifying regret as a factor in “qualitative decision analysis” which he referred to as the “chagrin factor” [35]. Djulbegovic and colleagues [8] propose two different regret based processes for specifying decision thresholds: 1) they revise the EUT model to incorporate regret, and 2) they also propose the concept of “acceptable regret” (i.e., described as the “level of regret which the decision maker can comfortably tolerate” [8]), which simplifies to a distinct decision threshold equation. In the case of dual processing theory, the concept of regret is incorporated into other theorical paradigm [12].

Hybrid theory

The hybrid theory is another common paradigm for non-EUT decision thresholds, discussed in 16 of 68 articles in our review (~ 24%). Although hybrid theory sometimes shares similar structures to EUT decision-thresholds, as it is often slightly modified version of EUT decision threshold [43], it merits separate consideration as a non-EUT decision threshold paradigm. What constitutes a “hybrid theory” is poorly defined despite relatively frequent use of this term. For example, Djulbegovic and colleagues [43] articulate a hybrid theory paradigm that incorporates a mechanism how concepts such as a patient’s “relative value” for specific outcomes can modify the expected value calculation (i.e., commonly incorporating utilities) to arrive at the optimal clinical decision [43]. Alternatively, other scholars have stated that, within the hybrid model, individuals are believed to “use a kind of intuitive threshold they cannot explicate but that is based on their knowledge and/or perception about the harm and benefit of a treatment” [51]. In broad terms, the literature describes the hybrid threshold model as retaining “the original EUT formulation but invites the decision-maker to weigh health outcomes (morbidities and mortalities) differently when they occur in patients with and without disease” [14].

Verp and Heckerling [41], the earliest proponents for a hybrid theory approach to decision thresholds (in the context of medical decision-making), describe a decision threshold approach that considers patients’ preferences for pre-natal testing [41]. Basinga and colleagues [51] provide an example of how a hybrid threshold model would calculate the associated decision threshold that incorporates a weighed value of morbidity with respect to mortality and weighed value of provoked death relative to natural death to incorporate a decision-maker’s preference. Within other hybrid theories, the element of patient choice is incorporated into the threshold equation [76, 79]. For example, Djulbegovic and colleagues [79] propose a modified threshold formulation where the quotient of absolute risk of a major adverse event occurring and relative risk ratio is multiplied by a constant that “refers to the patient’s relative value of avoiding treatment harms[…] with respect to the impact of disease without treatment” [79].

Dual processing theory

Six of the reviewed articles adopted a theoretical paradigm of dual processing or dual systems theory (hereafter, dual processing theory). Djulbegovic and colleagues [12] were the first to apply dual processing theory to medical decision thresholds. In this paradigm, it is presumed that people make decisions by drawing on a combination of “type I system” reasoning (i.e., affectively driven, fast, intuitive) and “type II system” reasoning (i.e., analytical, calculative, deliberative) [12]. Djulbegovic and colleagues [61] posit that physicians do use the threshold model to inform their decisions, and they claim that the dual processing model may best explain physicians’ decision thresholds [61].

Of the six articles that identify dual processing theory as a theoretical underpinning for decision thresholds, two explored dual processing theory from a theoretical or conceptual perspective using hypothetical vignettes to illustrate its potential application [12, 22], three articles were reviews [9, 14, 77], and one article reported on the explanatory power of dual processing theory relative to EUT and regret theory decision threshold models [61].

Information theory

Information theory is used infrequently in the medical decision-making literature and is most commonly used in fields such as economics or engineering [90]. We identified two articles that used information theory in relation to decision thresholds. Within information theory, uncertainty, or “entropy” as it is known in the field of thermodynamics [39], can be expressed in terms of the benefit to risk ratio of a particular therapy [39], effectively incorporating a measure of choice [39]. The articles by Asch and colleagues [36] and Djulbegovic and colleagues [39] theoretically incorporate information theory into the classic EUT model by Pauker and Kassirer [10, 11].

Signal detection theory

We identified four articles that referenced Signal Detection Theory (SDT) relative to medical decision thresholds. SDT extends the concept of decision thresholds by providing a mechanism to contrast “signal” or “hit” (i.e., true positives and false negatives) and “noise” or “miss” (i.e., false positives and true negatives) relative to a decision [13, 54, 58, 72].

Three of the four articles used SDT as a methodology to elicit an individual’s decision threshold when presented with hypothetical vignettes [13, 54, 58]. In the fourth article, Hozo and colleagues [72] addressed to the role of SDT from a theoretical perspective and identify the link between decision threshold models and SDT, fast-and-frugal decision trees (FFT) and evidence accumulation theory (EAT) [72].

Regression (Logistic/ Ordinal)

Our review identified 18 articles referencing a regression model for decision thresholds (~ 27%). Linear or logistic regression was a common method to estimate decision thresholds. Unlike many of other theories for informing decision thresholds in an ex-ante fashion, those thresholds that are identified through regression (commonly, logistic or ordinal) are almost exclusively decision thresholds that are determined ex-post.

Within articles that use a regression, Eisenberg and Hershey [31] are commonly cited for their four-step method for calculating the test and test-treatment thresholds within medical decision-making. Young and colleagues [33] build on Eisenberg and Hershey [31] and propose three different approaches to estimating the decision threshold: 1) modal distribution, 2) the unweighted midpoint, 2) the weighted midpoint. Plasencia and colleagues [38] elaborate on the logistic function as proposed by Hartz and colleagues [32], and reduce the dependencies on the approach proposed by Young and colleagues [33]. Specifically, Plasencia and colleagues [38], provide a regression paradigm that can account, “for other factors that may influence disease probability, or threshold, at which a particular proportion of physicians makes a decision” [38]. More recently, Ebell and colleagues [66] modify the logistic model proposed by Plasencia and colleagues [38] to estimate thresholds for both testing and treatment decisions. They explore testing and treatment thresholds for: influenza, acute coronary syndrome, pneumonia, deep vein thrombosis, and urinary tract infection [66]. Ebell and colleagues [66] are commonly cited for their use of online clinical vignettes, which (at the time) provided a novel way to explore test and treatment thresholds for common conditions.

Net-Benefit

We identified three articles that specifically explored the concept of decision thresholds through a net-benefit approach. Although commonly the net-benefit approach is synonymous with EUT, the articles included in this review adopt a modified net-benefit approach. For example, Glasziou and Irwig [40] propose a net-benefit approach to decision thresholds that incorporates a weighting function for specific patient values for specific outcomes. This approach shares similar properties to typical EUT models but does not explicitly reference EUT [40]. Vickers and colleagues [60, 70] explore a net-benefit approach relative to the selection of diagnostic tests and incorporate a decision threshold approach into the net-benefit equation alongside specificity and sensitivity.

Other theoretical paradigms

Our review identified several other theoretical paradigms (i.e., classified as “Other” in Table 2), however each of these paradigms was associated with just one publication and therefore is not discussed in greater detail here. These “other” approaches include linear information theory [36], info-gap theory [55], generalized linear receiver operator characteristic (GROC) curves [37], rituals [49], social judgement theory [13], general assessment and decision making model [13], systems dynamics [69], fast-and-frugal decision trees [72], evidence accumulation theory [72], therapeutic risk thresholds [74], and the smooth ambiguity model [85].

Contextual and subjective factors

Our review sought to identify what consideration an individual’s (patient or provider) subjective judgment(s) (e.g., attitudes, emotions, preferences, risk perceptions, values) have been given in a context of decision threshold(s) within medical decision-making (i.e., Objective 2).

Of the 95 articles included in the scoping review (Fig. 1), 44 consider contextual or subjective factors that may influence an individual’s decision threshold (~ 46%). Contextual and subjective factors were infrequently grounded within a theoretical paradigm for decision thresholds (24 out of 44 articles- ~ 55%). Of the articles that included theoretical paradigm, regression techniques were most used (11 out of 24 articles- ~ 46%) to determine decision thresholds in an ex-post fashion and to identify the factors associated with heterogeneous decision thresholds. Where theoretical paradigms were not cited or regression techniques were not employed, studies relied on a variety of parametric (e.g., t-tests, ANOVAs) and non-parametric (e.g., Wilcoxon singed rank test) tests to quantify the decision threshold.

Contextual factors

Two broad categories of contextual factors were discussed in the literature as potentially influencing an individual’s decision threshold: person-specific factors, and occupation-related factors (see Table 3). We categorize “person-specific factors” as those that provide contextual understanding of the individual (e.g., age, gender, country, education level, patient vs. physician). The studies with “person-specific factors” capture both physicians and lay individuals. Alternatively, the “occupation related factors” are specific to studies that have investigated the influence of various contextual factors relative to physician decision thresholds.

Table 3 Contextual Factors that may Influence Decision Threshold
Full size table

Across the identified articles there were mixed results regarding the statistical significance of several contextual factors associated with an individual’s decision threshold. Interpretation of these various contextual factors can be found in the supplementary text. We also highlight that aside from regression models, signal detection theory [13, 54, 58] and regret theory [64, 65, 86] were the most common theoretical paradigms sited relative to contextual factors influencing decision thresholds.

Subjective factors

While we identified several subjective factors that influence decision thresholds, their potential influence relative to decision thresholds was mentioned but not necessarily quantified. We categorize these factors in three broad categories: 1) health related factors, 2) personal factors, and 3) perceptive factors (see Table 4).

Table 4 Subjective Factors that may Influence Decision Thresholds
Full size table

Similar to the findings associated with contextual factors, the statistical significance associated with various subjective factors was inconsistent across studies. With a greater frequency than with contextual factors, studies discussed (but did not measure) the potential influence that subjective factors have on an individual’s decision threshold. Aside from regression models, we observe that hybrid theory was often used as a theoretical paradigm relative to subjective factors and decision thresholds (5 articles [41, 62, 67, 81, 85]). A plausible explanation for use of a hybrid theory being is that this theory incorporates decision-maker preferences and choices into the estimation of the decision threshold.

Discussion

Classification of theoretical paradigms: descriptive, normative and prescriptive

Theoretical models for decision-making can be classified as descriptive, normative, or prescriptive (see Table 5) [2, 110,111,112]. Descriptive and prescriptive approaches provide an alternative view of “rationality” compared to the normative approach which is most often consistent with EUT [113].

Table 5 Decision-Theoretical Classifications
Full size table

The decision-theoretical classification of theoretical paradigms provides guidance as to how the theoretical paradigms ought to be evaluated [114]. In our review, included studies often commented on the duality or plurality of possible classifications of their theoretical paradigm, emphasizing how the authors conceptualized their contribution and sought to evaluate their study. For example, Djulbegovic and colleagues (1999) discuss that the concept that “acceptable regret may have prescriptive as well as descriptive value” [8] and Asch and colleagues (1990) noted that their threshold models, informed by information theory, “are neither solely descriptive nor solely normative” [36]. While the regression method for decision thresholds may be commonly considered a descriptive decision-theoretical classification, it can also be considered normative owing to its connection to EUT. Dual processing theory has typically been characterized as a descriptive theory [12], yet hybrid theory, essentially a version of dual processing theory [14], has been previously classified as a prescriptive theory [51]. To further complicate efforts to categorize these models, not all authors provide reflections on/ identification of the decision-theoretical classification they associate with their research. Although classifying models and paradigms as descriptive, prescriptive, or normative could augment our understanding of their potential applications, accurately doing so will require future researchers to reflect and report on this in their publications.

Regret theory: a larger body of literature

We have identified regret theory as the most common non-EUT paradigm used to inform decision thresholds within the context of medical decision-making. Importantly, regret theory extends beyond the medical decision-making literature; regret research appears in many different fields, including economics, psychology, medicine, law, organizational behaviour, to name few [115].

Regret is often identified as a driver of ‘irrational’ decision-making given its tendency to influence people to make decisions that are inconsistent with EUT [116]. Under the traditional approach to decision-making, decisions are expected to be driven by EUT [117] where a ‘rational’ decision is one that maximizes the expected utility of the final assets [23, 24]. Other scholars maintain that because anticipated regret causes people to think more elaborately before making their decision, that regret can also induce rational decision-making [118].

Regret is a cognitively based emotion that refers to the affective reaction of unfavourable outcomes [119, 120]. Specifically, regret can be measured as “the difference between the utility of the action taken and the utility of the action that in retrospect should have been taken” [19]. Unlike other emotions, regret is uniquely tied to decision-making [115]. Regret can be experienced, ex-post, once the outcomes of a decision have become known, or can be anticipated, ex-ante, before the decision is made and is a reflection of how a person anticipates feeling if an undesirable outcome were to occur [118, 121]. Within medical decision-making most decisions cannot be reversed (e.g., surgery cannot be undone, a vaccine cannot be ungiven) and consequently decision-making is informed by anticipatory (ex-ante) regret [25]. Although decision thresholds are not specifically referenced, some of the early research on regret in medical decision-making was led by Ritov and Baron on vaccine hesitancy and omission bias [122, 123].

Regression: is it a novel theoretical paradigm?

Regression models are commonly employed to estimate decision thresholds in an ex-post fashion (i.e., after the decision has been made) in the medical decision-making literature. Within our scoping review we classified “regression” as a theoretical paradigm for decision thresholds to capture the articles that discuss decision thresholds and use regression models within the context of a medical decision. However, on a conceptual level, we maintain that “regression” should not necessarily be a considered a theoretical paradigm that is distinctive from EUT. To the extent that our inclusion and exclusion criteria were established a priori (as is appropriate for scoping review methodology) we could not exclude these articles because they do not explicitly reference an EUT theoretical paradigm These articles used regression as a method of analysis to estimate an individual’s decision threshold. To this end, the articles that leverage regression models explore various factors that may inform a decision threshold or use scales to derive a quantifiable decision threshold, but do not propose a novel theoretical construct (beyond EUT) to understand such a threshold.

Hybrid theory, net-benefit, information theory and personal preferences

Our review identified several different theoretical paradigms for decision thresholds. The theoretical paradigms for decision thresholds of: “hybrid theory”, “information theory” and “net-benefit” are distinctive in their own-right but often have similarities. Importantly, each of these paradigms are referred to as distinctive theoretical paradigms within the literature. However, within each of these theoretical paradigms the incorporation of the concept of “choice” or “values” was a common theme. To this end, while the underpinning mathematical equations to derive the associated decision threshold respective of each of these theoretical paradigms may be different, the consideration of an individual’s preferences (choices) differentiates these theoretical paradigms from those of regret theory, dual processing theory, signal detection theory and regression techniques.

Dual processing theory

Outside of the medical decision-making literature (e.g., sociology, cognition, etc.) the dual processing theory is, “widely accepted as a dominant explanation of cognitive processes that characterizes human decision-making” [12]. Within the fields of sociology and cognition the dual processing approach to decision-making is recognized as the dominant mechanism for reasoning [124,125,126,127]. Specifically, the dual processing theory adds a decisions threshold that is calculated through a combination of affective reasoning (i.e., “type I”) and analytical reasoning (i.e., “type II”) [12, 22]. Effectively, the DPM incorporates “type I system” reasoning by using anticipated regret as a proxy of the affect or emotion that is commonly used in “type I system” reasoning [12]. The DPM also incorporates “type II system” reasoning through an EUT approach [12]. It is the combination of “type I” and “type II” system reasoning to estimate a decision threshold that differentiates dual processing theory from regret theory and EUT. We, again, reflect that it has been claimed that the dual processing theory of decision thresholds has been demonstrated to be more consistent with physician decision-making than thresholds determined by EUT and regret theory [25]. It is also worthy of highlight that the hybrid theory is generally considered a version of the dual processing theory of decision-threshold analysis [14]. In this review we classified “dual processing theory” and “hybrid theory” as distinct theoretical paradigms for decision thresholds. However, if we accepted these theoretical paradigms as one and the same then dual processing theory would become the most frequently used theoretical paradigm for non-EUT decision thresholds within medical decision-making. It is important to note that in the literature dual processing theory and hybrid theory are recognized as distinct theoretical paradigms. Dual processing theory incorporates “type I” and “type II” system reasoning, whereas hybrid theory suggests that decision-makers should incorporate weighting/ preferences relative to the identified possible outcomes and does not consider a balance between “type I” and “type II” system reasoning. To this end, there may be the potential for hybrid theory to be incorporated into dual processing theory by way of the “type I” system reasoning.

Contextual and subjective factors: limited theoretical underpinning

This scoping review identified several contextual and subjective factors that have been reported in the literature to inform or influence an individual’s decision threshold. Importantly, the evidence on almost every contextual and subjective factor is inconsistent and the effects of the identified factors on decision thresholds is not certain. For example, five studies identified that age was a significant factor in determining an individual’s decision thresholds [34, 78, 82, 91, 92], whereas three other studies found age to not be a significant factor [93, 95, 128]. Part of the heterogeneity observed in the literature on the contextual and subjective factors influencing decision thresholds may be a result of poor theoretical conceptualization of decision thresholds. In the majority of studies where contextual or subjective factors were explored, there was little, if any, theorization of decision threshold and instead there was a reliance on regression models. Consequently, it may be possible that contextual and subjective factors have different levels of influence depending on how the decision threshold is theorized.

Contributions

The 2015 narrative review by Djulbegovic and colleagues [14] offered a comprehensive description of the research on decision thresholds in medical decision-making at that time. A decade later, we decided to re-examine the full extent of this research using a rigorous and systematic approach. Notably, the literature on decision thresholds in medical decision-making has matured and almost doubled in volume since 2013 – a limiting date used by Djulbegovic and colleagues in their review [14].

Djulbegovic and colleagues identified regret theory, hybrid theory, and dual processing theory as the theoretical paradigms for physician’s decision thresholds. While this finding remains true today, our review also revealed several other theoretical paradigms for decision thresholds were used by physicians and lay individuals. We also identified various contextual and subject factors that may influence an individual’s decision thresholds.

Limitations

Although our scoping review followed best methodological practices, it still has a few limitations. First, our review was limited to articles in English language that were indexed on SCOPUS, MEDLINE (Ovid), or ProQuest (Theses and Dissertations). Consequently, it is possible that our review failed to identify relevant articles that were not written in English and/or not indexed on either of the identified databases. Second, while our search strings and selection criteria were intentionally broad and used the most frequently employed terms, decision thresholds can be referred to by a wide variety of names (e.g., test threshold, diagnostic threshold, isolation/ quarantine threshold, thresholds for specific diseases, etc.). Thus, our review may have failed to identify relevant articles; however, we believe this is likely a small minority of articles which would not meaningfully change the overall findings of our systematic scoping review. Third, our review excluded decision thresholds that pertained to a clinical outcome assessment/ clinical guidelines/ prediction models (e.g., studies that sought to quantify the safety and accuracy of a diagnostic clinical aid) as well as policy thresholds (e.g., willingness-to-pay thresholds) as they are not informative of how individual’s make decisions. Fourth, within the included articles we reflect that there is a concentration of authors who are repeatedly included in the review. For example, Dr. Benjamin Djulbegovic is listed as an author on 21 of the 68 articles that discussed various theoretical paradigms for decision thresholds. While we do not perceive this to be a methodological limitation of our scoping review, the concentration of authors is an important consideration relative to the intellectual heterogeneity within the corpus of the literature. Fifth, our review did not attempt to interrogate the intricate differences between the different theoretical paradigms as it was beyond the scope of our review. It is possible that some of theoretical paradigms may share similar structures to EUT paradigms (e.g., regression, information theory, etc.). However, to the extent that these theoretical paradigms are uniquely identified within the literature, they warrant identification as non EUT theoretical paradigms within this review. Sixth, our review included articles that were indexed as of July 2023. Consequently, there may be new articles that were published since this time that are not captured in our review; ultimately, this is an unavoidable limitation of any systematic scoping review within an active and evolving field. Finally, our review had a relatively narrow focus on medical decision making. Given, decision thresholds are explored within other contexts (e.g., insurance [129,130,131], risk perception [132,133,134], law [3,4,5,6,7, 135] econometrics [110], etc.), it is possible that there may be additional theoretical paradigms and contextual/subjective factors (i.e., identified in other bodies of literature) than were discussed within this review.

Implications for practice and future research

From the perspective of clinical practice, decision thresholds have been called the “linchpin” between evidence based medicine and decision-making [2]. Yet, despite their significant role in clinical decision-making, there is a lack of consensus on which theoretical paradigm should be used to determine patient or physician decision thresholds [136]. Consequently, greater consideration should be given for choosing the theoretical paradigm used to inform ex-ante decision thresholds. From a practical perspective, if a physician can better understand a patient’s implicit decision threshold (e.g., perhaps through a regret theory or dual processing theory lens), they can improve patients’ decision-making by focusing on specific anticipated regrets of false positives or false negatives.

From a future research perspective, we have several recommendations. First, authors should be cognisant of the multiplicity of like-terms that are often used to reference decision thresholds (e.g., diagnostic thresholds, treatment thresholds, testing thresholds, etc.) which can only be resolved through a convergence within the academic community to adopt consistent terminology. We propose “decision thresholds” to be sufficiently flexible to capture the breadth and scope of related terms. Consistency in language allows for easier identification of literature and may help to avoid duplicative or redundant research effort. Secondly, we encourage authors to consider, and identify, within which decision-theoretical classification their paradigm should be interpreted (i.e., descriptively, normatively, and/or prescriptively). Third, we call for more research on how decision thresholds inform individual medical decisions. Specifically, it would be advantageous for future scholarship to conduct additional research on which theoretical paradigm for decision thresholds best explains an individual’s decision threshold for medical decisions. Consequently, this would help to narrow the number of theoretical paradigms that are used to inform decision threshold analyses [25, 71]. Finally, we encourage additional methodological research on the contextual and subjective factors that inform an individual’s decision threshold; additional research is required to better understand the impact these factors might have on decision thresholds.

Conclusions

To our knowledge, this is the first review of non-EUT decision thresholds used in medical decision-making that adopts a rigorous systematic search and reporting methodology (i.e., PRISMA-ScR). As the body of literature on the role of decision thresholds in medical decision-making continues to grow in popularity, this study offers a critical, systematic, characterization of the existing literature. Importantly, this study will help to ensure that authors of future scholarship to appropriately situate their work within the body of literature and leverage appropriate and relevant theoretical paradigms to underpin their understanding of decision thresholds.

Regret theory, hybrid theory, and dual processing theory were identified as the most common theoretical paradigms that are used to inform an individual’s ex-ante decision threshold, but other theories have been introduced in recent years. Further, although a substantial set of studies examine contextual and subjective factors that impact decision thresholds, we note considerable heterogeneity in the reported effect of these factors. We also observe a striking infrequency of theoretical grounding in these studies.

Availability of data and materials

The search strategies used to inform this scoping review are available in the supplementary text. Summaries of included articles are also available in the supplementary text.

References

  1. Brand KP, Finkel AM. A Decision-Analytic Approach to Addressing the Evidence About Football and Chronic Traumatic Encephalopathy. Semin Neurol. 2019;16:s-0039-1688484.

    Google Scholar 

  2. Djulbegović B, Hozo I. Threshold decision-making in clinical medicine : with practical application to hematology and oncology. Cham: Springer International Publishing AG; 2023. (Cancer Treatment and Research Series ; v.189); 2023.

    Book  Google Scholar 

  3. Dekay ML. The Difference between Blackstone-Like Error Ratios and Probabilistic Standards of Proof. Law Soc Inq. 1996;62:95–132.

    Article  Google Scholar 

  4. Weiss C. Expressing scientific uncertainty. Law Probab. Risk. 2003;2:25–46.

    Google Scholar 

  5. Weiss C. Scientific Uncertainty and Science-Based Precaution. Int Environ Agreem Polit Law Econ. 2003;3(2):137–66.

    Google Scholar 

  6. Dekay ML, Patiño-Echeverri D, Fischbeck PS. Better safe than sorry: Precautionary reasoning and implied dominance in risky decisions. J Behav Decis Mak. 2009;22(3):338–61.

    Article  Google Scholar 

  7. Dekay ML, Small MJ, Fischbeck PS, Farrow RS, Cullen A, Kadane JB, et al. Risk-based decision analysis in support of precautionary policies. J Risk Res. 2002;5(4):391–417.

    Article  Google Scholar 

  8. Djulbegovic B, Hozo I, Schwartz A, McMasters KM. Acceptable regret in medical decision making. Med Hypotheses. 1999;53(3):253–9.

    Article  CAS  PubMed  Google Scholar 

  9. Djulbegovic B, Hamm RM, Mayrhofer T, Hozo I, Van den Ende J. Rationality, practice variation and person-centred health policy: a threshold hypothesis. J Eval Clin Pract. 2015;21(6):1121–4.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Pauker S, Kassirer J. The threshold approach to clinical decision making. N Engl J Med. 1980;302(20):1109–17.

    Article  CAS  PubMed  Google Scholar 

  11. Stephen P, Jerome K. Therapeutic Decision Making: A Cost-Benefit Analysis. N Engl J Med. 1975;293(5):229–34.

    Article  Google Scholar 

  12. Djulbegovic B, Hozo I, Beckstead J, Tsalatsanis A, Pauker SG. Dual processing model of medical decision-making. BMC Med Inform Decis Mak. 2012;12(1):94.

    Article  PubMed  PubMed Central  Google Scholar 

  13. Cheyne H, Dalgleish L, Tucker J, Kane F, Shetty A, McLeod S, et al. Risk assessment and decision making about in-labour transfer from rural maternity care: a social judgment and signal detection analysis. BMC Med Inform Decis Mak. 2012;12(101088682):122.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Djulbegovic B, van den Ende J, Hamm RM, Mayrhofer T, Hozo I, Pauker SG, et al. When is rational to order a diagnostic test, or prescribe treatment: the threshold model as an explanation of practice variation. Eur J Clin Invest. 2015;45(5):485–93.

    Article  PubMed  Google Scholar 

  15. Djulbegovic B, Hozo I, Mayrhofer T, Ende J, Guyatt G. The threshold model revisited. J Eval Clin Pract. 2019;25(2):186–95.

    Article  PubMed  Google Scholar 

  16. Felder S, Mayrhofer T. Threshold analysis in the presence of both the diagnostic and the therapeutic risk. Eur J Health Econ. 2018;19(7):1019–26.

    Article  PubMed  Google Scholar 

  17. Djulbegovic B, Tsalatsanis A, Mhaskar R, Hozo I, Miladinovic B, Tuch H. Eliciting regret improves decision making at the end of life. Eur J Cancer. 2016;68:27–37.

    Article  PubMed  Google Scholar 

  18. Djulbegovic M, Beckstead J, Elqayam S, Reljic T, Kumar A, Paidas C, et al. Thinking Styles and Regret in Physicians. Antonietti A, editor. PLOS ONE. 2015;10(8):e0134038.

    Article  PubMed  PubMed Central  Google Scholar 

  19. Hozo I, Tsalatsanis A, Djulbegovic B. Expected utility versus expected regret theory versions of decision curve analysis do generate different results when treatment effects are taken into account. J Eval Clin Pract. 2018;24(1):65–71.

    Article  PubMed  Google Scholar 

  20. Hozo I, Djulbegovic B. When is diagnostic testing inappropriate or irrational? Acceptable regret approach. Med Decis Mak Int J Soc Med Decis Mak. 2008;28(4):540–53.

    Article  Google Scholar 

  21. Tsalatsanis A, Hozo I, Vickers A, Djulbegovic B. A regret theory approach to decision curve analysis: A novel method for eliciting decision makers’ preferences and decision-making. BMC Med Inform Decis Mak. 2010;10(51):1–14.

  22. Tsalatsanis A, Hozo I, Kumar A, Djulbegovic B. Dual Processing Model for Medical Decision-Making: An Extension to Diagnostic Testing. Brock G, editor. PLOS ONE. 2015;10(8):e0134800.

    Article  PubMed  PubMed Central  Google Scholar 

  23. Bell DE. Regret in Decision Making under Uncertainty. Oper Res. 1982;30(5):961–81.

    Article  Google Scholar 

  24. Loomes G, Sugden R. Regret Theory: An Alternative Theory of Rational Choice Under Uncertainty. Econ J. 1982;92(368):805.

    Article  Google Scholar 

  25. Djulbegovic B, Elqayam S, Reljic T, Hozo I, Miladinovic B, Tsalatsanis A, et al. How do physicians decide to treat: an empirical evaluation of the threshold model. BMC Med Inform Decis Mak. 2014;14(1):47.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Kahneman D, Tversky A. The Psychology of Preferences. Sci Am. 1982;246(1):160–73.

    Article  Google Scholar 

  27. Peters MDJ, Godfrey CM, Khalil H, McInerney P, Parker D, Soares CB. Guidance for conducting systematic scoping reviews. Int J Evid Based Healthc. 2015;13(3):141–6.

    Article  PubMed  Google Scholar 

  28. Scarffe AD, Coates A. Approaches to decision threshold models in medical decision-making: A scoping review protocol. OSF. 2023 Aug 29; Available from: https://osf.io/7dxr9/?view_only=.

  29. Tricco AC, Lillie E, Zarin W, O’Brien KK, Colquhoun H, Levac D, et al. PRISMA Extension for Scoping Reviews (PRISMA-ScR): Checklist and Explanation. Ann Intern Med. 2018 Oct 2 [cited 2023 Jul 29];169(7):467–73. Available from: https://www.acpjournals.org/doi/10.7326/M18-0850.

  30. Christensen-Szalanski JJ, Diehr PH, Bushyhead JB, Wood RW. Two studies of good clinical judgment. Med Decis Mak Int J Soc Med Decis Mak. 1982;2(3):275–83.

    Article  CAS  Google Scholar 

  31. Eisenberg JM, Hershey JC. Derived thresholds. Determining the diagnostic probabilities at which clinicians initiate testing and treatment. Med Decis Mak Int J Soc Med Decis Mak. 1983;3(2):155–68.

    Article  CAS  Google Scholar 

  32. Hartz A, McKinney WP, Centor R, Krieg A, Simms G, Henck S. Stochastic thresholds. Med Decis Mak Int J Soc Med Decis Mak. 1986;6(3):145–8.

    Article  CAS  Google Scholar 

  33. Young MJ, Eisenberg JM, Williams SV, Hershey JC. Comparing aggregate estimates of derived thresholds for clinical decisions. Health Serv Res. 1986;20(6 Pt 1):763–80.

    CAS  PubMed  PubMed Central  Google Scholar 

  34. Young M, Fried LS, Eisenberg J, Hershey J, Williams S. Do cardiologists have higher thresholds for recommending coronary arteriography than family physicians? Health Serv Res. 1987;22(5):623.

    CAS  PubMed  PubMed Central  Google Scholar 

  35. Feinstein AR. The “Chagrin Factor” and Qualitative Decision Analysis. Arch Intern Med. 1985;145(7):1257–9.

    Article  CAS  PubMed  Google Scholar 

  36. Asch DA, Patton JP, Hershey JC. Knowing for the sake of knowing: the value of prognostic information. Med Decis Mak Int J Soc Med Decis Mak. 1990;10(1):47–57.

    Article  CAS  Google Scholar 

  37. Sainfort F. Evaluation of medical technologies: a generalized ROC analysis. Med Decis Mak Int J Soc Med Decis Mak. 1991;11(3):208–20.

    Article  CAS  Google Scholar 

  38. Plasencia CM, Alderman BW, Baron AE, Rolfs RT, Boyko EJ. A method to describe physician decision thresholds and its application in examining the diagnosis of coronary artery disease based on exercise treadmill testing. Med Decis Mak Int J Soc Med Decis Mak. 1992;12(3):204–12.

    Article  CAS  Google Scholar 

  39. Djulbegovic B, Hozo I, Abdomerovic I, Hozo S. Diagnostic entropy as a function of therapeutic benefit/risk ratio. Med Hypotheses. 1995;45(5):503–9.

    Article  CAS  PubMed  Google Scholar 

  40. Glasziou PP, Irwig LM. An evidence based approach to individualising treatment. BMJ. 1995;311(7016):1356–9.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  41. Verp MS, Heckerling PS. Use of decision analysis to evaluate patients’ choices of diagnostic prenatal test. Am J Med Genet. 1995;58(4):337–44.

    Article  CAS  PubMed  Google Scholar 

  42. Hozo I, Djulbegovic B. Using the Internet to calculate clinical action thresholds. Comput Biomed Res Int J. 1999;32(2):168–85.

    Article  CAS  Google Scholar 

  43. Djulbegovic B, Hozo I, Lyman GH. Linking evidence-based medicine therapeutic summary measures to clinical decision analysis. MedGenMed Medscape Gen Med. 2000;2(1):E6.

    CAS  Google Scholar 

  44. Van Hoe L, Miserez M. Effectiveness of imaging studies in acute appendicitis: A simplified decision model. Eur J Emerg Med. 2000;7:25–40.

    Article  PubMed  Google Scholar 

  45. McAlister FA, O’Connor AM, Wells G, Grover SA, Laupacis A. When should hypertension be treated? The different perspectives of Canadian family physicians and patients. CMAJ Can Med Assoc J J Assoc Medicale Can. 2000;163(4):403–8.

    CAS  Google Scholar 

  46. Coenen S, Van Royen P, Vermeire E, Hermann I, Denekens J. Antibiotics for coughing in general practice: a qualitative decision analysis. Fam Pract. 2000;17(5):380–5.

    Article  CAS  PubMed  Google Scholar 

  47. Sinclair JC, Cook RJ, Guyatt GH, Pauker SG, Cook DJ. When should an effective treatment be used? Derivation of the threshold number needed to treat and the minimum event rate for treatment. J Clin Epidemiol. 2001;54(3):253–62.

    Article  CAS  PubMed  Google Scholar 

  48. Cotler SJ, Patil R, McNutt RA, Speroff T, Banaad-Omiotek G, Ganger DR, et al. Patients’ values for health states associated with hepatitis C and physicians’ estimates of those values. Am J Gastroenterol. 2001;96(9):2730–6.

    Article  CAS  PubMed  Google Scholar 

  49. Sonnenberg A. Personal view: cost and benefit of medical rituals in gastroenterology. Aliment Pharmacol Ther. 2004;20(9):939–42.

    Article  CAS  PubMed  Google Scholar 

  50. Ng AK, Li S, Neuberg D, Silver B, Weeks J, Mauch P. Factors influencing treatment recommendations in early-stage Hodgkin’s disease: a survey of physicians. Ann Oncol Off J Eur Soc Med Oncol. 2004;15(2):261–9.

    Article  CAS  Google Scholar 

  51. Basinga P, Moreira J, Bisoffi Z, Bisig B, Van den Ende J. Why Are Clinicians Reluctant to Treat Smear-Negative Tuberculosis? An Inquiry about Treatment Thresholds in Rwanda. Med Decis Making. 2007;27(1):53–60.

    Article  PubMed  Google Scholar 

  52. Hozo I, Schell MJ, Djulbegovic B. Decision-Making When Data and Inferences Are Not Conclusive: Risk-Benefit and Acceptable Regret Approach. Semin Hematol. 2008;45(3):150–9.

    Article  PubMed  Google Scholar 

  53. Moreira J, Bisig B, Muwawenimana P, Basinga P, Bisoffi Z, Haegeman F, et al. Weighing Harm in Therapeutic Decisions of Smear-Negative Pulmonary Tuberculosis. Med Decis Making. 2009;29(3):380–90.

    Article  PubMed  Google Scholar 

  54. Thompson C, Dalgleish L, Bucknall T, Estabrooks C, Hutchinson AM, Fraser K, et al. The effects of time pressure and experience on nurses’ risk assessment decisions: a signal detection analysis. Nurs Res. 2008;57(5):302–11.

    Article  PubMed  Google Scholar 

  55. Ben-Haim Y, Zacksenhouse M, Keren C, Dacso CC. Do we know how to set decision thresholds for diabetes? Med Hypotheses. 2009;73(2):189–93.

    Article  CAS  PubMed  Google Scholar 

  56. Boland MV, Lehmann HP. A new method for determining physician decision thresholds using empiric, uncertain recommendations. BMC Med Inform Decis Mak. 2010;10(101088682):20.

    Article  PubMed  PubMed Central  Google Scholar 

  57. Tsalatsanis A, Barnes LE, Hozo I, Djulbegovic B. Extensions to regret-based decision curve analysis: an application to hospice referral for terminal patients. BMC Med Inform Decis Mak. 2011;11(101088682):77.

    Article  PubMed  PubMed Central  Google Scholar 

  58. Mohan D, Rosengart MR, Farris C, Fischhoff B, Angus DC, Barnato AE. Sources of non-compliance with clinical practice guidelines in trauma triage: a decision science study. Implement Sci IS. 2012;7(101258411):103.

    Article  PubMed  Google Scholar 

  59. Pines JM, Lessler AL, Ward MJ, Mark CD. The mortality benefit threshold for patients with suspected pulmonary embolism. Acad Emerg Med. 2012;19(9):E1109–13.

    Article  PubMed  Google Scholar 

  60. Vickers AJ, Cronin AM, Gonen M. A simple decision analytic solution to the comparison of two binary diagnostic tests. Stat Med. 2013;32(11):1865–76.

    Article  PubMed  Google Scholar 

  61. Djulbegovic B, Elqayam S, Reljic T, Hozo I, Miladinovic B, Tsalatsanis A, et al. How do physicians decide to treat: an empirical evaluation of the threshold model. BMC Med Inform Decis Mak. 2014;14(101088682):47.

    Article  PubMed  PubMed Central  Google Scholar 

  62. Felder S, Mayrhofer T. Risk preferences: consequences for test and treatment thresholds and optimal cutoffs. Med Decis Mak Int J Soc Med Decis Mak. 2014;34(1):33–41.

    Article  Google Scholar 

  63. Hernandez JM, Tsalatsanis A, Humphries LA, Miladinovic B, Djulbegovic B, Velanovich V. Defining optimum treatment of patients with pancreatic adenocarcinoma using regret-based decision curve analysis. Ann Surg. 2014;259(6):1208–14.

    Article  PubMed  Google Scholar 

  64. Sreeramareddy CT, Rahman M, Harsha Kumar HN, Shah M, Hossain AM, Sayem MA, et al. Intuitive weights of harm for therapeutic decision making in smear-negative pulmonary Tuberculosis: an interview study of physicians in India, Pakistan and Bangladesh. BMC Med Inform Decis Mak. 2014;14(101088682):67.

    Article  PubMed  PubMed Central  Google Scholar 

  65. Cucchetti A, Djulbegovic B, Tsalatsanis A, Vitale A, Hozo I, Piscaglia F, et al. When to perform hepatic resection for intermediate-stage hepatocellular carcinoma. Hepatol Baltim Md. 2015;61(3):905–14.

    Article  Google Scholar 

  66. Ebell MH, Locatelli I, Senn N. A novel approach to the determination of clinical decision thresholds. Evid Based Med. 2015;20(2):41–7.

    Article  PubMed  Google Scholar 

  67. Sonnenberg A. Ignorance isn’t bliss: why patients become angry. Eur J Gastroenterol Hepatol. 2015;27(6):619–22.

    Article  PubMed  Google Scholar 

  68. Courbage C, Rey B. Decision Thresholds and Changes in Risk for Preventive Treatment. Health Econ. 2016;25(1):111–24.

    Article  PubMed  Google Scholar 

  69. Sheldrick RC, Breuer DJ, Hassan R, Chan K, Polk DE, Benneyan J. A system dynamics model of clinical decision thresholds for the detection of developmental-behavioral disorders. Implement Sci IS. 2016;11(1):156.

    Article  PubMed  Google Scholar 

  70. Vickers AJ, Van Calster B, Steyerberg EW. Net benefit approaches to the evaluation of prediction models, molecular markers, and diagnostic tests. BMJ Online. 2016;352. Available from: https://www.scopus.com/inward/record.uri?eid=2-s2.0-84959328652&doi=10.1136%2fbmj.i6&partnerID=40&md5=fe1ef082c28f9376c332111191d28ead.

  71. Tsalatsanis A, Hozo I, Djulbegovic B. Acceptable regret model in the end-of-life setting: Patients require high level of certainty before forgoing management recommendations. Eur J Cancer. 2017;75:159–66.

    Article  PubMed  Google Scholar 

  72. Hozo I, Djulbegovic B, Luan S, Tsalatsanis A, Gigerenzer G. Towards theory integration: Threshold model as a link between signal detection theory, fast-and-frugal trees and evidence accumulation theory. J Eval Clin Pract. 2017;23(1):49–65.

    Article  PubMed  Google Scholar 

  73. Ebell MH, Locatelli I, Mueller Y, Senn N, Morgan K. Diagnosis and treatment of community-acquired pneumonia in patients with acute cough: a quantitative study of decision thresholds in primary care. Br J Gen Pract. 2018;68(676):e765–74.

    Article  PubMed  PubMed Central  Google Scholar 

  74. Felder S, Mayrhofer T. Threshold analysis in the presence of both the diagnostic and the therapeutic risk. Eur J Health Econ. 2018;19(7):1019–26.

    Article  PubMed  Google Scholar 

  75. Fujii Y, Osaki Y. Regret-sensitive treatment decisions. Health. Econ Rev. 2018;8(1):14.

    Google Scholar 

  76. Djulbegovic B, Hozo I, Mayrhofer T, van den Ende J, Guyatt G. The threshold model revisited. J Eval Clin Pract. 2019;25(2):186–95.

    Article  PubMed  Google Scholar 

  77. Boyles TH, Lynen L, Seddon JA. Decision-making in the diagnosis of tuberculous meningitis. Tuberculous Meningitis International Research Consortium, editor. Wellcome Open Res. 2020;5(101696457):11.

    Article  PubMed  PubMed Central  Google Scholar 

  78. De Alencastro L, Locatelli I, Clair C, Ebell MH, Senn N. Correlation of clinical decision-making with probability of disease: A web-based study among general practitioners. PloS One. 2020;15(10 October):e0241210.

    Article  PubMed  PubMed Central  Google Scholar 

  79. Djulbegovic M, Lee AI, Chen K. Which patients with unprovoked venous thromboembolism should receive extended anticoagulation with direct oral anticoagulants? A systematic review, network meta-analysis, and decision analysis. J Eval Clin Pract. 2020;26(1):7–17.

    Article  PubMed  Google Scholar 

  80. Patel BS, Steinberg E, Pfohl SR, Shah NH. Learning decision thresholds for risk stratification models from aggregate clinician behavior. J Am Med Inform Assoc JAMIA. 2021;28(10):2258–64.

    Article  PubMed  Google Scholar 

  81. Courbage C, Peter R. On the effect of uncertainty on personal vaccination decisions. Health Econ. 2021;30(11):2937–42.

    Article  PubMed  PubMed Central  Google Scholar 

  82. van Overbeeke E, Hauber B, Michelsen S, Peerlinck K, Lambert C, Hermans C, et al. Patient preferences for gene therapy in haemophilia: Results from the PAVING threshold technique survey. Haemophilia. 2021;27(6):957–66.

    Article  PubMed  PubMed Central  Google Scholar 

  83. Cai X, Ebell MH, Geyer RE, Thompson M, Gentile NL, Lutz B. The impact of a rapid home test on telehealth decision-making for influenza: a clinical vignette study. BMC Prim Care. 2022;23(1):75.

    Article  PubMed  PubMed Central  Google Scholar 

  84. Wen FH, Chou WC, Hou MM, Su PJ, Shen WC, Chen JS, et al. Associations of surrogates’ death-preparedness states with decisional conflict and heightened decisional regret over cancer patients’ last 6 months of life. Psychooncology. 2022;31(9):1502–9.

    Article  PubMed  Google Scholar 

  85. Sevim D, Felder S. Decision Thresholds for Medical Tests Under Ambiguity Aversion. Front Health Serv. 2022;2(9918334887706676):825315.

    Article  PubMed  PubMed Central  Google Scholar 

  86. Cucchetti A, Djulbegovic B, Crippa S, Hozo I, Sbrancia M, Tsalatsanis A, et al. Regret affects the choice between neoadjuvant therapy and upfront surgery for potentially resectable pancreatic cancer. Reg-PanC study group, editor. Surg U S. 2023;173(6):1421–7.

    Google Scholar 

  87. Djulbegovic B, Hozo I, Lizarraga D, Guyatt G. Decomposing clinical practice guidelines panels’ deliberation into decision theoretical constructs. J Eval Clin Pract. 2023;29(3):459–71.

    Article  PubMed  Google Scholar 

  88. Taylor SP, Weissman GE, Kowalkowski M, Admon AJ, Skewes S, Xia Y, et al. A Quantitative Study of Decision Thresholds for Initiation of Antibiotics in Suspected Sepsis. Med Decis Mak. 2023;43(2):175–82.

    Article  Google Scholar 

  89. Bell DE. Risk Premiums for Decision Regret. Manag Sci. 1983;29(10):1156–66.

    Article  Google Scholar 

  90. Diamond GA, Hirsch M, Forrester JS, Staniloff HM, Vas R, Halpern SW, et al. Application of information theory to clinical diagnostic testing. The electrocardiographic stress test Circulation. 1981;63(4):915–21.

    CAS  PubMed  Google Scholar 

  91. Steel N. Thresholds for taking antihypertensive drugs in different professional and lay groups: questionnaire survey. BMJ. 2000;320(7247):1446–7.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  92. Brundage MD, Feldman-Stewart D, Cosby R, Gregg R, Dixon P, Youssef Y, et al. Cancer patients’ attitudes toward treatment options for advanced non-small cell lung cancer: implications for patient education and decision support. Patient Educ Couns. 2001;45(2):149–57.

    Article  CAS  PubMed  Google Scholar 

  93. Cahan A, Gilon D, Manor O, Paltiel O. Probabilistic reasoning and clinical decision-making: do doctors overestimate diagnostic probabilities? QJM Mon J Assoc Physicians. 2003;96(10):763–9.

    Article  CAS  Google Scholar 

  94. Douglas H. Weighing Complex Evidence in a Democratic Society. Kennedy Inst Ethics J. 2012;22(2):139–62.

    Article  PubMed  Google Scholar 

  95. Nair T, Savulescu J, Everett J, Tonkens R, Wilkinson D. Settling for second best: when should doctors agree to parental demands for suboptimal medical treatment? J Med Ethics. 2017;43(12):831–40.

    Article  PubMed  Google Scholar 

  96. Lahaye S, Regpala S, Lacombe S, Sharma M, Gibbens S, Ball D, et al. Evaluation of patients’ attitudes towards stroke prevention and bleeding risk in atrial fibrillation. Thromb Haemost. 2014;111(3):465–73.

    CAS  PubMed  Google Scholar 

  97. van der Keylen P, Zeschick N, Schlenz AR, Kuhlein T. Treatment thresholds and minimal clinically important effect sizes of antiosteoporotic medication-Survey among physicians and lay persons in Germany. PloS One. 2022;17(8 August):e0272985.

    Article  PubMed  PubMed Central  Google Scholar 

  98. Winkenwerder W, Levy BD, Eisenberg JM, Williams SV, Young MJ, Hershey JC. Variation in physicians’ decision-making thresholds in management of a sexually transmitted disease. J Gen Intern Med. 1993;8(7):369–73.

    Article  CAS  PubMed  Google Scholar 

  99. Hanson LC, Danis M, Garrett JM, Mutran E. Who decides? Physicians’ willingness to use life-sustaining treatment. Arch Intern Med. 1996;156(7):785–9.

    Article  CAS  PubMed  Google Scholar 

  100. Connors GR, Siner JM. Clinical Reasoning and Risk in the Intensive Care Unit. Clin Chest Med. 2015;36(3):449–59.

    Article  PubMed  Google Scholar 

  101. Di Stefano LM, Wood K, Mactier H, Bates SE, Wilkinson D. Viability and thresholds for treatment of extremely preterm infants: survey of UK neonatal professionals. Arch Dis Child Fetal Neonatal Ed. 2021;106(6):F596–602.

    Article  Google Scholar 

  102. Stojan JN, Daniel M, Hartley S, Gruppen L. Dealing with uncertainty in clinical reasoning: A threshold model and the roles of experience and task framing. Med Educ. 2022;56(2):195–201.

    Article  PubMed  Google Scholar 

  103. Greenfield S, Bryan S, Gill P, Gutridge K, Marshall T. Factors influencing clinicians’ decisions to prescribe medication to prevent coronary heart disease. J Clin Pharm Ther. 2005;30(1):77–84.

    Article  CAS  PubMed  Google Scholar 

  104. Ost DE, Gould MK. Decision making in patients with pulmonary nodules. Am J Respir Crit Care Med. 2012;185(4):363–72.

    Article  PubMed  PubMed Central  Google Scholar 

  105. Donner-Banzhoff N, Muller B, Beyer M, Haasenritter J, Seifart C. Thresholds, rules and defensive strategies: how physicians learn from their prior diagnosis-related experiences. Diagn Berl Ger. 2020;7(2):115–21.

    Article  Google Scholar 

  106. Billington EO, Feasel AL, Kline GA. At Odds About the Odds: Women’s Choices to Accept Osteoporosis Medications Do Not Closely Agree with Physician-Set Treatment Thresholds. J Gen Intern Med. 2020;35(1):276–82.

    Article  PubMed  Google Scholar 

  107. Man-Son-Hing M, Gage BF, Montgomery AA, Howitt A, Thomson R, Devereaux PJ, et al. Preference-based antithrombotic therapy in atrial fibrillation: implications for clinical decision making. Med Decis Mak Int J Soc Med Decis Mak. 2005;25(5):548–59.

    Article  Google Scholar 

  108. Minami CA, King TA, Mittendorf EA. Patient preferences for locoregional therapy in early-stage breast cancer. Breast Cancer Res Treat. 2020;183(2):291–309.

    Article  PubMed  Google Scholar 

  109. Lazarus DR, Ost DE. The solitary pulmonary nodule-deciding when to act? Semin Respir Crit Care Med. 2013;34(6):748–61.

    Article  PubMed  Google Scholar 

  110. Kahneman D, Tversky A. Prospect Theory: An Analysis of Decision under Risk. Econometrica. 1979;47(2):263–92.

    Article  Google Scholar 

  111. Baron J. Normative, descriptive and prescriptive responses. Behav Brain Sci. 1994;17(1):32–42.

    Article  Google Scholar 

  112. Baron J. The point of normative models in judgment and decision making. Front Psychol. 2012 [cited 2024 Aug 18];3. Available from: http://journal.frontiersin.org/article/10.3389/fpsyg.2012.00577/abstract.

  113. Djulbegovic B, Elqayam S. Many faces of rationality: Implications of the great rationality debate for clinical decision-making. J Eval Clin Pract. 2017;23(5):915–22.

    Article  PubMed  PubMed Central  Google Scholar 

  114. Bell DE, Raiffa H, Tversky A. Descriptive, normative, and prescriptive interactions in decision making. In: Bell DE, Raiffa H, Tversky A, editors. Decision Making. 1st ed. Cambridge University Press; 1988 [cited 2024 Aug 18]. p. 9–30. Available from: https://www.cambridge.org/core/product/identifier/CBO9780511598951A010/type/book_part.

  115. Zeelenberg M, Pieters RA. Theory of Regret Regulation 1.0. J Consum Psychol. 2007;17(1):3–18.

    Article  Google Scholar 

  116. Sugden R. Regret, recrimination and rationality. Theory Decis. 1985;19:77–99.

    Article  Google Scholar 

  117. Von Neumann J, Morgenstern O. Theory of games and economic behavior, 2nd rev. 1947.

    Google Scholar 

  118. Zeelenberg M. Anticipated regret, expected feedback and behavioral decision making. J Behav Decis Mak. 1999;12(2):93–106.

    Article  Google Scholar 

  119. Landman J. Regret and Elation Following Action and Inaction: Affective Responses to Positive Versus Negative Outcomes. Pers Soc Psychol Bull. 1987;13(4):524–36.

    Article  Google Scholar 

  120. Zeelenberg M, Pieters R. Consequences of regret aversion in real life: The case of the Dutch postcode lottery. Organ Behav Hum Decis Process. 2004;93(2):155–68.

    Article  Google Scholar 

  121. Zeelenberg M, Beattie J, van der Pligt J, de Vries NK. Consequences of Regret Aversion: Effects of Expected Feedback on Risky Decision Making. Organ Behav Hum Decis Process. 1996;65(2):148–58.

    Article  Google Scholar 

  122. Ritov I, Baron J. Reluctance to vaccinate: Omission bias and ambiguity. J Behav Decis Mak. 1990;3:263–77.

    Article  Google Scholar 

  123. Ritov I, Baron J. Outcome Knowledge, Regret, and Omission Bias. Organ Behav Hum Decis Process. 1995;64(2):119–27.

    Article  Google Scholar 

  124. Kahneman D. Thinking, Fast and Slow, vol. 1st. Straus and Giroux: Edition. Farrar; 2011.

    Google Scholar 

  125. Lizardo O, Mowry R, Sepulvado B, Stoltz DS, Taylor MA, Van Ness J, et al. What Are Dual Process Models? Implications for Cultural Analysis in Sociology. Sociol Theory. 2016;34(4):287–310.

    Article  Google Scholar 

  126. Strandell J. Bridging the Vocabularies of Dual-Process Models of Culture and Cognition. In: Brekhus WH, Ignatow G, editors. The Oxford Handbook of Cogntive Sociology. 2019.

    Google Scholar 

  127. Vaisey S. Motivation and Justification: A Dual-Process Model of Culture in Action. Am J Sociol. 2009;114(6):1675–715.

    Article  Google Scholar 

  128. Douglas F, Petrie KJ, Cundy T, Horne A, Gamble G, Grey A. Differing perceptions of intervention thresholds for fracture risk: a survey of patients and doctors. Osteoporos Int. 2012;23(8):2135–40.

    Article  CAS  PubMed  Google Scholar 

  129. Robinson PJ, Botzen WJW. The impact of regret and worry on the threshold level of concern for flood insurance demand: Evidence from Dutch homeowners. Judgm Decis Mak. 2018;13(3):237–45.

    Article  Google Scholar 

  130. Robinson PJ, Botzen WJW. Determinants of Probability Neglect and Risk Attitudes for Disaster Risk: An Online Experimental Study of Flood Insurance Demand among Homeowners. Risk Anal. 2019;39(11):2514–27.

    Article  PubMed  PubMed Central  Google Scholar 

  131. Robinson PJ, Botzen WJW. Flood insurance demand and probability weighting: The influences of regret, worry, locus of control and the threshold of concern heuristic. Water Resour Econ. 2020;30: 100144.

    Article  Google Scholar 

  132. Slovic P, Fischhoff B, Lichtenstein S, Corrigan B, Combs B. Preference for Insuring against Probable Small Losses: Insurance Implications. J Risk Insur. 1977;44(2):237.

    Article  Google Scholar 

  133. Botzen WJW, Kunreuther H, Michel-Kerjan E. Divergence between individual perceptions and objective indicators of tail risks: Evidence from floodplain residents in New York City. Judgm Decis Mak. 2015;10(4):21.

    Article  Google Scholar 

  134. Botzen WJW, Duijndam SJ, Robinson PJ, Beukering P. Behavioral biases and heuristics in perceptions of COVID‐19 risks and prevention decisions. Risk Anal. 2022;29:risa.13882.

    Google Scholar 

  135. Lempert RO. Modeling Relevance. Mich Law Rev. 1977;75(5/6):1021.

    Article  Google Scholar 

  136. Djulbegovic B, Hozo I, Mandrola J. Sorites paradox and persistence in overuse and underuse in healthcare delivery services. J Eval Clin Pract. 2023;29(6):877–9.

    Article  PubMed  Google Scholar 

Download references

Acknowledgements

This manuscript is based on research presented in a chapter of AS’s doctoral dissertation at the University of Ottawa. The authors would also like to thank the anonymous peer-reviewers of this manuscript whose intellectual reflections contributed positively to this paper.

Funding

AS would like to acknowledge the financial support of the following scholarships: Mitacs Accelerate Entrepreneur Graduate Scholarship, the Ontario Graduate Scholarship, the Queen Elizabeth II Graduate Scholarship for Science and Technology, and the University of Ottawa. AC would like to acknowledge the financial support from the following scholarships: Canada Graduate Scholarship -Doctoral (CGS D) from the Social Sciences and Humanities Research Council (SSHRC), and the University of Ottawa. KB was the primary investigator for the Mitacs Accelerate Entrepreneur Graduate Scholarship. All funding agreements ensured the authors’ independence in designing the study, interpreting the data, writing, and publishing this manuscript.

Author information

Authors and Affiliations

  1. Telfer School of Management, University of Ottawa, Ottawa, ON, Canada

    Andrew Scarffe, Alison Coates, Kevin Brand & Wojtek Michalowski

  2. Ottawa Hospital Research Institute, The Ottawa Hospital, Ottawa, ON, Canada

    Andrew Scarffe

  3. Bob Gaglardi School of Business and Economics, Thompson Rivers University, Kamloops, BC, Canada

    Andrew Scarffe

Authors
  1. Andrew Scarffe
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Alison Coates
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Kevin Brand
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Wojtek Michalowski
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

Study concept and design: AS, AC, KB, WM; acquisition of data: AS, AC; critical revision of manuscript: AS, AC, KB, WM; final approval of the version to be published: AS, AC, KB, WM; agreement to be accountable for all aspects of the work: AS, AC, KB, WM.

Corresponding author

Correspondence to Andrew Scarffe.

Ethics declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Scarffe, A., Coates, A., Brand, K. et al. Decision threshold models in medical decision making: a scoping literature review. BMC Med Inform Decis Mak 24, 273 (2024). https://doi.org/10.1186/s12911-024-02681-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12911-024-02681-2

Keywords

BMC Medical Informatics and Decision Making

ISSN: 1472-6947

Contact us