Five Different Types of Framing Effects in Medical Situation: A Preliminary Exploration

Jiaxi Peng, Wanqing Li, Danmin Miao, Xi Feng, Wei XiaoIranian 2013 · DOI

TL;DR

This study explores how different ways of presenting medical information (framing effects) influence decision-making. It found that even logically equivalent descriptions can lead to significantly different choices, highlighting the need for standardized communication in healthcare to ensure patients and doctors make informed decisions.

What Problem Does This Paper Address?

The paper addresses the pervasive phenomenon of "framing effects" in medical decision-making. Framing effects occur when the way information is presented, rather than the information itself, alters people's choices. While this effect is well-documented in general decision-making (e.g., the classic Asian Disease Problem), this study specifically investigates different types of framing effects within medical contexts and compares their impact. The authors aim to understand how variations in framing, such as focusing on positive outcomes (lives saved) versus negative outcomes (lives lost), or presenting information in terms of survival rates versus mortality rates, influence patient and doctor choices. This is crucial because medical decisions often involve high stakes, and understanding these cognitive biases is essential for improving patient care and adherence to treatments.

Methods

The researchers employed a quantitative experimental design. They created five distinct medical decision-making scenarios, each framed in both a positive and a negative way. These scenarios covered:

  • Attribute Framing: Describing a product's quality (e.g., "80% lean" vs. "20% fat").
  • Goal Framing: Emphasizing the benefits of an action versus the consequences of inaction.
  • Risky Choice Framing (Options Equivalent): Presenting choices with identical outcomes but different wording (e.g., lives saved vs. lives lost).
  • Risky Choice Framing (Options Not Equivalent): Presenting choices where the outcomes are not perfectly symmetrical but are framed differently (e.g., survival rates vs. mortality rates).
  • Number-Size Framing: Leveraging the concept of diminishing marginal utility, where people are more sensitive to smaller numerical changes than larger ones.

Participants (500 male undergraduates) were randomly assigned to one of ten groups, with each group tackling a single framed decision-making problem. Responses were collected using a 6-point Likert scale, allowing for the measurement of preference strength. The study then analyzed the effect sizes of these different framing types using Cohen's d to compare their magnitude.

Datasets

Not discussed in the paper. The study used custom-designed hypothetical medical decision-making scenarios rather than pre-existing datasets.

Key Insights and Ideas

  • Framing is Pervasive in Medicine: The study confirms that framing effects are not just theoretical curiosities but significantly influence medical choices, affecting evaluations of drug efficacy, patient compliance with doctor's advice, and treatment selection.
  • Different Frames, Different Decisions: Even when the underlying information is logically identical, the way it's presented can lead to opposite preferences. For instance, a positive frame might encourage risk aversion, while a negative frame might promote risk-seeking behavior.
  • Magnitude Matters: Not all framing effects are created equal. The study found varying effect sizes across the five types, with "number-size framing" showing the strongest impact, suggesting that how numerical information is presented can be particularly influential.
  • Prospect Theory as an Explanatory Framework: The authors link their findings to Prospect Theory, explaining that people's decisions are influenced by perceived gains and losses relative to a reference point, and that losses are generally felt more acutely than equivalent gains.

Key Results

  • Attribute Framing: Participants evaluated a drug more positively when described as "70 patients became better" compared to "30 patients didn't become better" (P < 0.01).
  • Goal Framing: Doctor's advice framed in terms of benefits of compliance led to better adherence than advice framed in terms of negative consequences of non-compliance (P < 0.05).
  • Risky Choice Framing (Equivalent): In the classic Asian Disease Problem, participants were more conservative in the positive frame (lives saved) and more risk-seeking in the negative frame (lives lost) (P < 0.01).
  • Risky Choice Framing (Non-Equivalent): When presented with survival rates, participants favored riskier options, whereas mortality rates led to more conservative choices (P < 0.05).
  • Number-Size Framing: This type of framing also had a significant effect (P < 0.01).
  • Effect Sizes: The study found that goal framing had a small effect size (Cohen's d = 0.40), attribute and risky choice framing (equivalent and non-equivalent) had medium effect sizes (0.64, 0.52, 0.42 respectively), and number-size framing had a large effect size (1.21).

Limitations and Future Work

The authors acknowledge that their study used hypothetical scenarios and a specific participant demographic (male undergraduates). They suggest that future research should explore these framing effects in real-world clinical settings with diverse patient populations and healthcare professionals. Additionally, further investigation into the nuances of each framing type and their interaction with individual differences could provide deeper insights. The primary recommendation for practice is to standardize how medical information is presented to minimize the influence of framing effects and ensure more objective decision-making.

Summary updated

Abstract

Medical decision making can be affected by frame descriptions. Attentions should be paid on the standardization of description in medical practice.
BibTeX
@inproceedings{peng2013five,
  author = {Jiaxi Peng and Wanqing Li and Danmin Miao and Xi Feng and Wei Xiao},
  title = {{Five Different Types of Framing Effects in Medical Situation: A Preliminary Exploration}},
  booktitle = {Iranian},
  year = {2013},
  doi = {10.5812/ircmj.8469},
}
Figure