CRD summary

The review investigated whether narrative testimonials can affect decision making and if so what factors mediated this. The authors found limited evidence to suggest that narrative affected decision making, but it remained unclear whether this biased decision making. Limitations of the methodology and reporting of the review made the reliability of the results questionable.

Authors' objectives

To determine what factors might explain the influence of testimonial and/or narrative information on patients' decision making.

Searching

MEDLINE (1966 to August 2007), PsycINFO (1806 to August 2007) and Communication Abstracts database (1977 to August 2007) were searched. Journal of Behavioural Decision Making, Medical Education Journal and Medical Decision Making were handsearched. Search dates spanned 1995 to August 2007. Bibliographies of identified studies, prior reviews and an unpublished doctoral thesis were searched, an author search was performed using Google Scholar and key authors were contacted. Only English-language studies were eligible for inclusion.

Study selection

Studies with the primary aim of measuring the role of first-person narrative (an account of an individual’s experience, often called testimonial evidence) or third-person narrative (an account of an individual’s experience conveyed in the third person) evidence in health decision making and that required adult participants to make a real or hypothetical decision or behaviour were eligible for inclusion. Design inclusion criteria were experimental design and/or randomised controlled trials (RCTs) and before/after or cohort studies that compared narrative information to data in another format (such as statistical evidence).

Most included studies were experimental; some were follow-up surveys. The comparison arm was provision of no additional information and/or statistic-based information. All but one study was carried out in North America. Participants were either student or general population samples. Most studies included third-person narratives and presented hypothetical decisions scenarios. The outcomes assessed were individual preferences and/or decision making.

Abstracts and titles were evaluated for eligibility. Full manuscripts were retrieved for abstracts that met the inclusion criteria, had no abstract and in which the abstract information was insufficient to make a judgment about inclusion.

It was unclear how many reviewers performed the study selection, but the authors stated that 20% of decisions about the inclusion and exclusion of articles were discussed between all authors and revisions were made to the inclusion and exclusion criteria as necessary.

Assessment of study quality

A validity assessment was conducted. A scoring system based on seven methodological or reporting aspects of the studies was used. The criteria were: sample size; sampling procedure; timing; validation of measures; consistency between study aims, methods, results and conclusions; and causes concern. The total research quality score was coded as 1 for yes, 0 for unsure or not relevant and -1 for no, with a resulting possible range of -7 to +7.

The authors did not state how many reviewers performed the validity assessment.

Data extraction

The authors developed a form for data abstraction, but stated neither how data were extracted for the review nor how many reviewers performed the data extraction.

Methods of synthesis

Studies were synthesised using descriptive statistics and qualitatively. The studies were grouped by narrative type (first- or third-person narrative).

Results of the review

Seventeen studies (n=3,986) were included in the review: 41% included first-person narratives and 59% included third-person narratives. The mean research quality rating was 3.5 (range 2 to 4); this did not differ significantly by narrative type. Most studies had an adequate sample size, but sampling was poorly described in all studies. Only one study included validated measures.

First person narratives: Seven studies covered the following scenarios (where specified): skin cancer prevention, testicular self-examination and polio vaccination. Three studies provided evidence that first-person narrative evidence had a persuasive effect on individual's decision. Three studies found that statistical and narrative data were equally effective.

Third-person narratives: Ten studies covered the following scenarios (where specified): organ donation, osteoporosis, skin cancer susceptibility. Two studies found that narrative was more persuasive than statistical information. Seven studies found both narrative and statistical messages to be persuasive under certain conditions.

Variables that were investigated to possibly explain the effect of narrative information on decision making were presented.

Authors' conclusions

Although it seemed likely that narrative information affected individuals' decision making processes and outcomes, it was unclear why narratives affected the decision making process, whether they facilitated or biased decision making and whether they affected the quality and/or outcome of the decision.

CRD commentary

The search question was broad but clear. The search strategy was extensive and attempts were made to identify unpublished data. The restriction to English-language articles meant that the results may have been affected by language bias. It was unclear how many reviewers performed the study selection and data extraction, so errors or bias may have been introduced during the review process. A validity assessment was conducted, although reporting of this indicated that it may have been rather subjective and no details were given on how it was performed. The narrative synthesis was appropriate.

Some aspects of the review process are not well reported, but the authors'conclusions are suitably cautious and reflect the limited evidence presented.

Implications of the review for practice and research

Practice: The authors suggested that people who designing interventions to facilitate informed decision making avoid use of patient testimonials until there was evidence to explain what type of narrative encouraged bias in information processing and decision making and which mechanisms mediated the effect.

Research: The authors stated that studies should aim to strengthen validity by using non-student or patient populationa and/or scenariosinformed by current health practice. The quality of research could also be improved by explicit reporting of the sampling procedure and inclusion of validated measures. Further research is also needed to examine the impact of including different narrative types in hypothetical scenarios.

Funding

Financial support for this study was provided partly by a PhD studentship (Economic and Social Research Council and Baxter’s Pharmaceuticals).

Bibliographic details

Winterbottom A, Bekker HL, Conner M, Mooney A. Does narrative information bias individual's decision making? A systematic review Social Science and Medicine 2008; 67(12): 2079-2088. [PubMed: 18951673]

Original Paper URL

http://dx.doi.org/doi:10.1016/j.socscimed.2008.09.037

Indexing Status

Subject indexing assigned by NLM

MeSH

Decision Making; Evidence-Based Medicine; Humans; Narration; Patient Participation; Persuasive Communication

AccessionNumber

12009101733

Database entry date

24/02/2010

Record Status

This is a critical abstract of a systematic review that meets the criteria for inclusion on DARE. Each critical abstract contains a brief summary of the review methods, results and conclusions followed by a detailed critical assessment on the reliability of the review and the conclusions drawn.