Unit 5. Quality Assurance and Quality Improvement
Quality Assurance & Quality Improvement

Section 4. Data Collection and Analysis

Developing a Patient-Satisfaction Survey

To develop a patient-satisfaction survey, consider assembling a focus group of key staff members, patients, and other community members. Through group processes such as brainstorming, nominal group process, and multi-voting, the focus group can arrive at a group of topics that are of concern to the staff, administration, and clinic users. Survey questions can then be developed from the list of topics that the focus group agreed on. Once the survey instrument has been completed, the focus group can also be used to pilot-test the instrument to ensure that it is gathering the information the group wants to collect. 

There are few scientifically validated patient-satisfaction surveys available for ambulatory health care in general or for dentistry in particular. Every clinic and service population will have unique problems and concerns, so you will probably need to develop your own survey instrument.

Constructing questions for your survey using a Likert scale can be valuable if the selection options are clear. A Likert scale measures the extent to which a person agrees or disagrees with a statement. The most common scale is 1 to 5; for example, 1=strongly disagree, 2=disagree, 3=not sure, 4=agree, and 5=strongly agree. A frequency distribution of the responses for each item is a simple and effective method for analyzing responses to Likert scales.

A five-point Likert scale would look something like the following:

Please answer the following questions by marking the appropriate column.

Strongly Agree

Agree

Neutral

Disagree

Strongly Disagree

Does Not Apply

Tips for Formatting and Wording a Survey

  • Consider the average educational level of the population for which the survey is intended. Questions should be understandable to the majority of those who will fill the survey out, without being condescending.
  • If your service population is bilingual or multi-lingual, translate the survey into the appropriate languages, without introducing misunderstanding or bias.
  • Word all questions clearly and concisely.
  • Word questions neutrally so as to not lead the respondent to a desired answer or otherwise introduce bias.
  • Simple yes/no or true/false questions are easily interpreted and easy to analyze but fail to identify areas in the middle.
  • Many patient-satisfaction surveys use the five-point or seven-point Likert scale to attempt to quantify the middle ground.
  • Multiple choice responses and written answers can also be used, but they tend to be more difficult to analyze statistically.
  • Ask for comments and suggestions at the end of the survey; respondents are likely to point out areas of concern or satisfaction.
  • Keep the survey short and simple (one page or less whenever possible); your user population will be more likely to participate if it does not take too much time.
  • Use a font of at least 12–14 points to make text easy to read.

Resource

  • Patient-satisfaction survey examples are listed in Section 3.