Empirical relationships between numeracy and treatment decision making: A scoping review of the literature

Ethical dilemmas surrounding patients´ "unwise" treatment preferences and suboptimal decision quality: case series of three renal cell carcinoma patients who developed local recurrences after non-guideline-concordant care choices

Patient engagement and shared decision-making (SDM) between patients and clinicians is the foundation of patient-centered care. It aims to reach a treatment option that fits the patient's preference and is guideline-concordant. We sought to evaluate the possible causes and outcomes of patient's non-guideline-concordant care choices. Using a retrospective analysis of the medical records of patients who underwent cryoablation for small renal masses between January 2010 and January 2023. Inclusion criteria were patients with renal tumor(s) who underwent cryoablation which was not recommended by the multidisciplinary team (MDT). We present three patients with unilateral clear cell renal cell carcinoma. Based on imaging and other findings, the oncology MDT recommended partial/radical nephrectomy. Upon consultation, each refused surgery and preferred cryoablation. Respecting their choice, cryoablation was undertaken. The patients had treatment failure and developed recurrences that could have possibly been avoided with guideline-concordant care. Shared decision-making in healthcare involves several aspects: patient/family; uncertainty of available evidence of various treatments; MDT meetings; and treatment team. For patients to select 'wise' treatment preferences i.e. guideline-concordant care, multi-layered complex intellectual and cognitive processes are required, where experience may play a role. Healthcare professionals require guidance and training on appropriate SDM in clinical settings, and awareness of tools to solicit patient choice to guideline-concordant care whilst observing patient autonomy. Patients and treatment teams need the capacity, knowledge, and skills to reach a 'wise' guideline-concordant care treatment preference jointly. Patients' unwise preference could lead to suboptimal outcomes, in the case of our patients, tumor recurrence.

What Constitutes High Risk for Venous Thromboembolism? Comparing Approaches to Determining an Appropriate Threshold

Background

Guidelines recommend pharmacological venous thromboembolism (VTE) prophylaxis only for high-risk patients, but the probability of VTE considered "high-risk" is not specified. Our objective was to define an appropriate probability threshold (or range) for VTE risk stratification and corresponding prophylaxis in medical inpatients.

Methods

Patients were adults admitted to any of 10 Cleveland Clinic Health System hospitals between December 2020 and August 2021 (N = 41,036). Hospital medicine physicians and internal medicine residents from included hospitals were surveyed between June and November 2023 (N = 214). We compared five approaches to determining a threshold: decision analysis, maximizing the sensitivity and specificity of a logistic regression model, deriving a probability from a point-based model, surveying physicians' understanding of VTE risk, and deriving a probability from physician behavior. For each approach, we determined the probability threshold above which a patient would be considered high-risk for VTE. We applied each threshold to the Cleveland Clinic VTE risk assessment model (CCM) and calculated the percentage of the 41,036 patients in our cohort who would be considered eligible for prophylaxis due to their high-risk status. We compared these hypothetical prophylaxis rates with physicians' observed prophylaxis rates.

Results

The different approaches yielded thresholds ranging from 0.3% to 5.4%, corresponding inversely with hypothetical prophylaxis rates of 0.2% to 75%. Multiple thresholds clustered between 0.52% to 0.55%, suggesting an average hypothetical prophylaxis rate of approximately 30%, whereas physicians' observed prophylaxis rates ranged from 48% to 76%.

Conclusions

Multiple approaches to determining a probability threshold for VTE prophylaxis converged to suggest an optimal threshold of approximately 0.5%. Other approaches yielded extreme thresholds that are unrealistic for clinical practice. Physicians prescribed prophylaxis much more frequently than the suggested rate of 30%, indicating opportunity to reduce unnecessary prophylaxis. To aid in these efforts, guidelines should explicitly quantify high-risk.

Numeracy levels influence shared decision-making and surgical outcomes: A scoping review of the literature

BACKGROUND

Health literacy has been widely studied with regard to medical decision-making and health care access, however research regarding numeracy - the ability to comprehend and attach meaning to numbers - is more limited.

METHODS

A scoping review following PRISMA guidelines was conducted. We screened 132 abstracts and 12 studies were included in the analysis.

RESULTS

Surgical population numeracy ranged from 47 to 86.1%. We found heterogeneity in the scales used to measure numeracy and the cutoff values used to define adequate numeracy. Low numeracy was shown to influence the accuracy of patients' responses to quality of life measures used to determine surgical outcomes and was associated with patient overestimation of pre-operative risk. Adequate numeracy was correlated with improved outcomes 2-4 years after bariatric surgery.

CONCLUSIONS

Patient numeracy is generally poor and has important implications for pre-operative risk understanding, accuracy of health measurement tools and long-term surgical outcomes.

Talking about Risk, UncertaintieS of Testing IN Genetics (TRUSTING): development and evaluation of an educational programme for healthcare professionals about BRCA1 & BRCA2 testing

BACKGROUND

Mainstreaming of germline testing demands that all healthcare professionals have good communication skills, but few have genetic testing and counselling experience. We developed and evaluated educational workshops-Talking about Risk & UncertaintieS of Testing IN Genetics (TRUSTING). Contents included: presentations and exercises, an interview with a geneticist about BRCA testing, screening and prevention implications, filmed interactions between surgeons, a genetic counsellor and geneticists with a fictitious family (proband had a BRCA2 pathogenic variant with triple-negative breast cancer, her older sister-BRCA2 heterozygous, and cousin-negative for BRCA2 variant).

METHODS

Twenty-one surgeons, 5 oncologists, 18 nurses and 9 genetic counsellors participated. Knowledge (18 item MCQ), communication skills (responses to 6 questions from proband and relatives) and self-confidence (discussing 9 genetic testing issues) were assessed pre- and post workshop.

RESULTS

Knowledge scores improved significantly post workshop (mean change = 7.06; 95% confidence interval (CI) 6.37-7.74; P < 0.001), as did communication (mean change = 5.38; 95% CI 4.37-6.38; P < 0.001) and self-confidence (P < 0.001).

DISCUSSION

Healthcare professionals' knowledge and self-confidence when discussing the risks and uncertainties in genetics are often poor. TRUSTING workshops significantly enhanced attendees' navigation of communication difficulties encountered and will be rolled out more widely.

The role of emotional sensitivity to probability in the decision to choose genetic testing

While the availability of genetic testing is rapidly increasing, many opt out of testing. The decision to test or not is emotionally charged, and both clinical research and theoretical work in psychology show that in emotional decisions, people often struggle to interpret and utilize risk information. Clinical research on genetic testing uptake also shows that feeling overwhelmed by numeric information may be a deterrent to testing. However, recent psychological research indicates that some portion of the population has greater emotional sensitivity to probability, (i.e., the extent to which emotional reactions to risk depend on probabilities) than others. We hypothesize that participants high in emotional sensitivity to probability will be more interested in genetic testing as an opportunity to seek greater precision in risk estimates and that this relationship is moderated by the testing context itself. In an online survey of a nationally representative sample (not recruited as patients), participants were presented with a hypothetical scenario describing a suspected diagnosis with an option for genetic testing. The scenario experimentally varied the pre-test probability estimate of the diagnosis (low or high) and whether the test results would result in certainty (ruling in or ruling out the diagnosis), or reduced uncertainty (providing a more precise individual risk estimate). Results indicated that emotional sensitivity to probability was a strong predictor of intention to test, particularly when the test allowed participants to rule out a diagnosis, reducing an already low probability to zero. These results highlight the way patients' individual characteristics interact with the testing context to guide decision-making and provide important insight into the way patients integrate risk probability information into intensely emotional decisions.

Pairing facts with imagined consequences improves pandemic-related risk perception

The COVID-19 pandemic reached staggering new peaks during a global resurgence more than a year after the crisis began. Although public health guidelines initially helped to slow the spread of disease, widespread pandemic fatigue and prolonged harm to financial stability and mental well-being contributed to this resurgence. In the late stage of the pandemic, it became clear that new interventions were needed to support long-term behavior change. Here, we examined subjective perceived risk about COVID-19 and the relationship between perceived risk and engagement in risky behaviors. In study 1 (n = 303), we found that subjective perceived risk was likely inaccurate but predicted compliance with public health guidelines. In study 2 (n = 735), we developed a multifaceted intervention designed to realign perceived risk with actual risk. Participants completed an episodic simulation task; we expected that imagining a COVID-related scenario would increase the salience of risk information and enhance behavior change. Immediately following the episodic simulation, participants completed a risk estimation task with individualized feedback about local viral prevalence. We found that information prediction error, a measure of surprise, drove beneficial change in perceived risk and willingness to engage in risky activities. Imagining a COVID-related scenario beforehand enhanced the effect of prediction error on learning. Importantly, our intervention produced lasting effects that persisted after a 1- to 3-wk delay. Overall, we describe a fast and feasible online intervention that effectively changed beliefs and intentions about risky behaviors.

Impact of numeracy preferences on information needs for genome sequencing results

OBJECTIVE

This study investigated how self-reported numeracy ability and preferences predict preferences for the amount and types of information provided about genome sequencing results among 1080 women diagnosed with breast cancer at age 40 or younger.

METHODS

Participants reported their level of interest in 14 topics related to genome sequencing results on a survey. We calculated a Participant Information Needs (PIN) value based on the number of topics for which a participant wanted "a lot" of information. Numeracy was assessed using the Subjective Numeracy Scale. Analyses examined associations between the numeracy ability and preferences subscales, information needs for individual content topics, and PIN.

RESULTS

Higher preference for numeric data was correlated with increased PIN (β = 0.60, p < 0.01), while numeric ability was not correlated (β=0.16, p = 0.22). Family composition and knowledge about sequencing benefits were also significant covariates. Patients most preferred information on topics related to disease risk and health implications.

CONCLUSION

There may be utility in separating numeracy ability and preferences into two components in future research in order to investigate how numeracy impacts the return of genetic testing results.

PRACTICE IMPLICATIONS

These data suggest that numeracy preferences may be important to inform strategies for the return of genetic results.

Factors predicting collaborative willingness of surrogates making medical decisions on the Physician Order for Scope of Treatment (POST)

Objective: The Physician's Order for Scope of Treatment (POST) indicates patient preferences regarding cardiopulmonary resuscitation (CPR), levels of care, and fluids/nutrition provision decisions. When patients become incapacitated, 'surrogates' often collaborate with physicians on POST decisions. Surrogates may vary in their willingness to collaborate, which can be problematic when physicians expect shared decision-making. No research has yet investigated collaborative decision-making among surrogates on the POST. This study investigated how six psychological variables predicted participants' desires for collaboration when completing an online decision-making task.Methods: Participants served as hypothetical surrogates and made decisions for another person on the three sections of the West Virginia POST. One-hundred-and-seventy-two adults were recruited from Amazon Mechanical Turk.Results: The six variables contributed significantly to the prediction of collaborative willingness, F (6, 163) = 5.29, p < .001, R²= 0.19. Two variables uniquely contributed: confidence and consideration of future consequences. The model most strongly predicted collaborative willingness for the CPR decision.Conclusion: This study provides a novel examination of under-researched areas: surrogate collaborative willingness and the POST. Differing risks associated with the three POST decisions may influence how surrogates value collaboration. Ways to enhance collaborative willingness when making POST decisions are discussed.

Framing the numerical findings of Cochrane plain language summaries: two randomized controlled trials

Background

Cochrane systematic review Plain language Summaries (CSR PLSs should serve as a tool for the evidence translation to non-medical population. However, the evidence of optimal type of numerical presentation in CSR PLSs is still scarce. The aim of this study was to investigate readers’ comprehension and preferences for different presentation of findings, including framing and numerical data, in Cochrane systematic review Plain Language Summaries (CSR PLSs).

Methods

We conducted a parallel randomized trial and a crossover randomized trial at the School of Medicine and family practice offices in Split, Croatia. The participants were students and consumers. We assessed possible differences in comprehension, measured by four questions on PLS content, of CSR PLSs depending on the positive or negative framing of results (n = 91) (Trial 1) or using percentages or frequencies for the presentation of results (n = 245) (Trial 2). The outcome measures were comprehension of PLS content, perceived effectiveness of the treatment and readiness to use the treatment (all on 1–10 scales).

Results

In Trial 1 we found no difference in readers’ perception of the effectiveness of the described treatment, desire that the treatment be offered by their family doctor, readiness to use the treatment, or comprehension when CSR PLS results were presented positively or negatively. In Trial 2 we found no difference in CSR PLS comprehension when results were presented as natural frequencies or percentages (BF₁₀ = 0.62, Bayesian t-test for independent samples).

Conclusions

Numerical presentation and framing direction of results appear to have no significant impact on understanding of messages in CSR PLSs.

Trial registration

The trials were registered in ClinicalTrials.gov. Protocol registration numbers: Trial 1: NCT03442387; Trial 2: NCT03554252.

The numerical translation of verbal probability expressions by patients and clinicians in the context of peri-operative risk communication

Shared decision-making is central to the pre-operative consent process and accurate communication of risk is dependent on a clear understanding of numerical information by both the patient and clinician. The risk of an adverse event or complication is often described using verbal probability expressions but how these are interpreted by clinicians and patients in the pre-operative setting has not been studied. We asked patients and clinicians to assign a numerical translation (as a percentage) for seven verbal probability expressions in relation to the probability of a major peri-operative complication occurring. In total, data from 290 patients and 57 clinicians were analysed. There was a wide range in percentages assigned by patients to all verbal probability expressions. Patients assigned a wider range of percentage values to each of the verbal probability expressions and these were all significantly higher than those assigned by clinicians: median (IQR [range]) negligible risk 5% (1-15 [0-100]) vs. 0% (0-0 [0-5]); minimal risk 5% (2-10) [0-100]) vs. 1% (0-1 [0-10]); low risk 10% (3-10 [0-100]) vs. 1% (0-2) [0-10]); standard risk 20% (10-40) [0-100]) vs. 1% (1-5) [0-30]); moderate risk 33% (20-50) [0-100]) vs. 5% (3-10) [0-80]); high risk 70% (30-90 [0-100]) vs. 15% (10-40) [1-75]); and very high risk 90% (50-95 [0-100]) vs. 40% (20-50 [5-100]), respectively (p < 0.005 for all comparisons). There is considerable variation in the numerical translation of verbal probability expressions by both patients and clinicians. This suggests that verbal probability expressions should not be used in isolation as part of doctor-patient discussions regarding peri-operative risk.

Health numeracy skills of medical students:cross-sectional and controlled before-and-after study

BACKGROUND

Although numeracy, defined as understanding and handling numbers, is an important skill for the medical profession, it is not clear whether it changes during graduate medical education and whether it can be improved by specific interventions. The objective of this study was to assess objective and subjective numeracy levels at different stages of medical education and explore whether a research methodology/statistics course improves numeracy levels in a longer period.

METHODS

We performed cross-sectional and controlled before-and-after studies. First-year sociology students and first- to sixth-year medical students from the in the cross sectional study and two groups of first-year medical students in a controlled before-and-after study. The intervention was a course on biostatistics and research methodology using blended approach. Numeracy was measured using Subjective Numeracy Scale (Cronbach α = 0.70) and Numeracy Understanding in Medicine instrument (Cronbach α = 0.75).

RESULTS

Whereas first-year medical students did not differ from first-year sociology students in objective numeracy, medicine students had higher results on subjective numeracy. Students from higher years of medical school had generally higher subjective and objective numeracy scores. In the controlled before-and-after study, the intervention group improved more in subjective numeracy (median difference on a 0-8 scale = 0.5, 95% CI 0.3 to 0.7 vs - 0.4, 95% CI - 0.4 to - 0.1, P < 0.001) but not in objective numeracy.

CONCLUSIONS

Although the numeracy levels at the beginning of the medical school are within the range of non-medical population, both objective and subjective numeracy improve during the higher years of medical school. Curriculum during medical school may help in numeracy increase, while research methodology training may help to increase subjective but not objective numeracy skills.

Health Coaching Has Differential Effects on Veterans with Limited Health Literacy and Numeracy: a Secondary Analysis of ACTIVATE

BACKGROUND

Health coaching is an effective behavior change strategy. Understanding if there is a differential impact of health coaching on patients with low health literacy has not been well investigated.

OBJECTIVE

To determine whether a telephone coaching intervention would result in similar improvements in enrollment in prevention programs and patient activation among Veterans with low versus high health literacy (specifically, reading literacy and numeracy).

DESIGN

Secondary analysis of a randomized controlled trial.

PARTICIPANTS

Four hundred seventeen Veterans with at least one modifiable risk factor: current smoker, BMI ≥ 30, or < 150 min of moderate physical activity weekly.

METHODS

A single-item assessment of health literacy and a subjective numeracy scale were assessed at baseline. A logistic regression and general linear longitudinal models were used to examine the differential impact of the intervention compared to control on enrollment in prevention programs and changes in patient activation measures (PAM) scores among patients with low versus high health literacy.

RESULTS

The coaching intervention resulted in higher enrollment in prevention programs and improvements in PAM scores compared to usual care regardless of baseline health literacy. The coaching intervention had a greater effect on the probability of enrollment in prevention programs for patients with low numeracy (intervention vs control difference of 0.31, 95% CI 0.18, 0.45) as compared to those with high numeracy (0.13, 95% CI - 0.01, 0.27); the low compared to high differential effect was clinically, but not statistically significant (0.18, 95% CI - 0.01, 0.38; p = 0.07). Among patients with high numeracy, the intervention group had greater increases in PAM as compared to the control group at 6 months (mean difference in improvement 4.8; 95% CI 1.7, 7.9; p = 0.003). This led to a clinically and statistically significant differential intervention effect for low vs high numeracy (- 4.6; 95% CI - 9.1, - 0.15; p = 0.04).

CONCLUSIONS

We suggest that health coaching may be particularly beneficial in behavior change strategies in populations with low numeracy when interpretation of health risk information is part of the intervention. CLINICALTRIALS.

GOV IDENTIFIER

NCT01828567.

Divide and conquer: improving parental understanding of health-related instructions using sequential pictorial instructions

Patient misunderstandings of healthcare-related instructions result in significant health, social and economic consequences. Pushing the boundaries of traditional academic methods may offer strategies that promote new ways to improve communication and understanding between healthcare providers, caregivers and patients. Comics can be an educational tool, offering an innovative strategy to communicate health-related information. Our study objective was to compare parents' understanding of health-related instructions (i.e. prescription labels, medical instructions and emergency situation scenarios) that were presented using two different formats-sequential pictorial instructions (SPIs) versus written text instructions (WRIs).This was a cross-sectional study to compare parents' understanding of health-related instructions for pediatric patients using structured SPIs versus WRIs. Parents (n = 359, 18-82 years old) of pediatric patients were given a card with instructions that were presented as SPIs (n = 195, 54.3%) or WRIs (n = 164, 45.7%), and then they completed a questionnaire. Total (14 questions) and grouped scores (10 questions) for parental comprehension of prescription and medical emergency instructions were significantly higher among participants viewing the SPIs than those viewing the WRIs in both English and Spanish. The mean understanding of instructions significantly increased from 6% to 12.2% when comparing the WRI versus the SPI.Our initial findings indicate that parents and caregivers of pediatric patients had significantly better understanding and recall of instructions delivered using SPIs than using WRIs.

Emotional processes in risky and multiattribute health decisions

OBJECTIVE

Here, we develop an integrative account of the roles of emotion in decision-making. In Part I, we illustrate how emotional inputs into decisions may rely on physiological signals from emotions experienced while making the decision, and we review evidence suggesting that the failure to represent the emotional meaning of options can often reduce decision quality. We propose that health-related decrements in the ability to generate emotional reactions lead people to inaccurately represent emotional responses and compromise decisions, particularly about risk. Part II explores complex decisions in which choice options involve trade-offs between positive and negative attributes. We first review evidence showing that difficult trade-off decisions generate negative affect and physiological arousal. Next, we propose that medical decision-making will be linked to short- and long-term stress and health outcomes.

CONCLUSION

In sum, this article proposes and reviews initial evidence supporting the effective use and management of emotional inputs as important to both clinical and non-clinical populations. Our approach will contribute to the understanding of patient-centred emotional decision-making and will inform medical decision aids.

Role of Psychosocial Factors and Health Literacy in Pregnant Women's Intention to Use a Decision Aid for Down Syndrome Screening: A Theory-Based Web Survey

BACKGROUND

Deciding about undergoing prenatal screening is difficult, as it entails risks, potential loss and regrets, and challenges to personal values. Shared decision making and decision aids (DAs) can help pregnant women give informed and values-based consent or refusal to prenatal screening, but little is known about factors influencing the use of DAs.

OBJECTIVE

The objective of this study was to identify the influence of psychosocial factors on pregnant women's intention to use a DA for prenatal screening for Down syndrome (DS). We also added health literacy variables to explore their influence on pregnant women's intention.

METHODS

We conducted a survey of pregnant women in the province of Quebec (Canada) using a Web panel. Eligibility criteria included age >18 years, >16 weeks pregnant, low-risk pregnancy, and having decided about prenatal screening for the current pregnancy. We collected data based on an extended version of the Theory of Planned Behavior assessing 7 psychosocial constructs (intention, attitude, anticipated regret, subjective norm, descriptive norm, moral norm, and perceived control), 3 related sets of beliefs (behavioral, normative, and control beliefs), 4 health literacy variables, and sociodemographics. Eligible women watched a video depicting the behavior of interest before completing a Web-based questionnaire. We performed descriptive, bivariate, and ordinal logistic regression analyses.

RESULTS

Of the 383 eligible pregnant women who agreed to participate, 350 pregnant women completed the Web-based questionnaire and 346 were retained for analysis (completion rate 350/383, 91.4%; mean age 30.1, SD 4.3, years). In order of importance, factors influencing intention to use a DA for prenatal screening for DS were attitude (odds ratio, OR, 9.16, 95% CI 4.02-20.85), moral norm (OR 7.97, 95% CI 4.49-14.14), descriptive norm (OR 2.83, 95% CI 1.63-4.92), and anticipated regret (OR 2.43, 95% CI 1.71-3.46). Specific attitudinal beliefs significantly related to intention were that using a DA would reassure them (OR 2.55, 95% CI 1.73-4.01), facilitate their reflections with their spouse (OR 1.55, 95% CI 1.05-2.29), and let them know about the advantages of doing or not doing the test (OR 1.53, 95% CI 1.05-2.24). Health literacy did not add to the predictive power of our model (P values range .43-.92).

CONCLUSIONS

Implementation interventions targeting the use of a DA for prenatal screening for DS by pregnant women should address a number of modifiable factors, especially by introducing the advantages of using the DA (attitude), informing pregnant women that they might regret not using it (anticipated regret), and presenting the use of DAs as a common practice (descriptive norm). However, interventions on moral norms related to the use of DA should be treated with caution. Further studies that include populations with low health literacy are needed before decisive claims can be made.

Are cancer-related decision aids appropriate for socially disadvantaged patients? A systematic review of US randomized controlled trials

BACKGROUND

Shared decision-making (SDM) is considered a key component of high quality cancer care and may be supported by patient decision aids (PtDAs). Many patients, however, face multiple social disadvantages that may influence their ability to fully participate in SDM or to use PtDAs; additionally, these social disadvantages are among the determinants of health associated with greater cancer risk, unwarranted variations in care and worse outcomes. The purpose of this systematic review is to describe the extent to which disadvantaged social groups in the United States (US) have been included in trials of cancer-related PtDAs and to highlight strategies, lessons learned and future opportunities for developing and evaluating PtDAs that are appropriate for disadvantaged populations.

METHODS

We selected cancer-related US studies from the Cochrane 2014 review of PtDAs and added RCTs meeting Cochrane criteria from searches of PubMed, CINAHL, PsycINFO (January 2010 to December 2013); and reference lists. Two reviewers independently screened titles/abstracts; three reviewers independently screened full text articles, performed data extraction and assessed: 1) inclusion of participants based on seven indicators of social disadvantage (limited education; female gender; uninsured or Medicaid status; non-U.S. nativity; non-White race or Hispanic ethnicity; limited English proficiency; low-literacy), and 2) attention to social disadvantage in the development or evaluation of PtDAs.

RESULTS

Twenty-three of 39 eligible RCTs included participants from at least one disadvantaged subgroup, most frequently racial/ethnic minorities or individuals with limited education and/or low-literacy. Seventeen studies discussed strategies and lessons learned in attending to the needs of disadvantaged social groups in PtDA development; 14 studies targeted disadvantaged groups or addressed subgroup differences in PtDA evaluation.

CONCLUSIONS

The diversity of the US population is represented in a majority of cancer-related PtDA RCTs, but fewer studies have tailored PtDAs to address the multiple social disadvantages that may impact patients' participation in SDM. More detailed attention to the comprehensive range of social factors that determine cancer risk, variations in care and outcomes is needed in the development and evaluation of PtDAs for disadvantaged populations.

TRIAL REGISTRATION

Registered 24 October 2014 in PROSPERO International prospective register of systematic reviews ( CRD42014014470 ).