Survey Design Principles for Quality Assessment

Surveys are widely used tools for capturing patient experiences, staff perceptions, and other subjective data in healthcare quality assessment. The CAHPS family of surveys is perhaps the most well known, but quality professionals may need to design or evaluate custom surveys as well. Understanding survey design principles helps ensure the data you collect is meaningful and actionable.

Question Construction

Effective survey questions are clear, concise, and focused on a single concept. Avoid double-barreled questions that ask about two things at once, such as "Was your nurse friendly and knowledgeable?" A respondent who found the nurse friendly but not knowledgeable cannot answer accurately. Use simple language that matches your respondents' reading level. Avoid jargon, acronyms, and technical terms unless your audience is specifically trained. Questions should be neutral; leading questions that suggest a desired answer will bias your results.

Response Scales and Options

Choose response scales that match your measurement goals. Likert scales (strongly disagree to strongly agree) work well for measuring attitudes and opinions. Frequency scales (never, sometimes, usually, always) suit behavioral questions. Include enough response options to capture meaningful variation, typically 4 to 7 points. Decide in advance whether to include a neutral midpoint. Always include a "not applicable" option when some questions may not apply to all respondents. Avoid forcing responses that do not reflect the respondent's true experience.

Reducing Bias in Surveys

Several types of bias can compromise survey results. Social desirability bias occurs when respondents answer in ways they believe are more acceptable rather than truthful. Acquiescence bias leads respondents to agree with statements regardless of content. Recency bias causes respondents to overweight recent experiences. To reduce these biases, mix positively and negatively worded items, ensure anonymity when possible, and survey close to the time of the experience being assessed.

Survey Administration

The method of administration affects response rates and data quality. Mail surveys offer convenience but may have low response rates. Telephone surveys allow for clarification but are more expensive. Electronic surveys are cost-effective but may exclude populations without internet access. In-person surveys have high response rates but may introduce interviewer bias. Consider your target population and resources when choosing a method.

Analyzing and Acting on Survey Results

Analyze survey data at levels that allow action. Organization-wide averages can mask important unit-level variation. Track results over time to identify trends. Compare against benchmarks when available. Most importantly, close the feedback loop by communicating results to staff and patients, and by developing improvement plans based on findings. A survey that collects data without driving improvement is a wasted effort.