Translation Guidelines for the Surveys on Patient Safety Culture

These guidelines were developed to help translate the Agency for Healthcare Research and Quality (AHRQ) Surveys on Patient Safety Culture. The 7-step process outlined here uses a team approach based on current best practices for survey translations.

These guidelines were developed to help translate the Agency for Healthcare Research and Quality (AHRQ) Surveys on Patient Safety Culture. The 7-step process outlined here uses a team approach based on current best practices for survey translations.



Prepared by Westat under contract number HHSA 290200710024C for the Agency for Healthcare Research and Quality (AHRQ).

Contents

Purpose and Use of This Document
Existing Translations
Step 1: Select Your Translation Team
Step 2: Develop a Draft Translation
Step 3: Review Draft Translation
Step 4: Conduct Initial Review and Signoff
Step 5: Pretest the Translation
Step 6: Conduct Final Review and Signoff
Step 7 (Optional): Conduct Field Test and Psychometric Analyses
References
Appendix A
Appendix B

Purpose and Use of This Document

These guidelines were developed to help you translate one of the Agency for Healthcare Research and Quality (AHRQ) Surveys on Patient Safety Culture (i.e., the Hospital, Medical Office, or Nursing Home Surveys). The 7-step process outlined here uses a team approach based on current best practices for survey translations (e.g., Douglas and Craig, 2007; Forsyth, et al., 2007; Institute of Survey Research, 2008; U.S. Census Bureau, 2009; Weidmer et al., 2006). Following these guidelines will help you achieve your translation goals—that is, to ensure that the survey's items, instructions, and responses:

  1. Convey the same meanings as the English version.
  2. Use language that is familiar to and easily understood by respondents.
  3. Are culturally appropriate.
  4. Use correct grammar and syntax.

Return to Contents

Existing Translations

The Hospital Survey on Patient Safety Culture and the Nursing Home Survey on Patient Safety Culture are both available in Spanish. The Spanish translations are designed for U.S. Spanish-speaking respondents from different countries. A number of translations in other languages have already been developed by international users who have agreed to share their translations. To obtain a copy of another translation, you can E-mail safetyculturesurveys@ahrq.hhs.gov (subject line: International Translations) and indicate the language you are interested in.

Return to Contents

Step 1: Select Your Translation Team

During the translation process, team members should talk to one another frequently. A basic team should include people to fill the following roles:

  • A translation coordinator to manage all translation activities, including identifying and recruiting other translation team members, providing them with background materials and information, and coordinating activities and communication among them. The coordinator should be familiar with the AHRQ Survey on Patient Safety Culture, understand the meaning and intent of the survey items and be able to convey the translation goals.
  • A translator fluent in English and the target language to produce a first draft translation for review and pretesting. Because the translator's task is to accurately convey the intended meaning of the English version, rather than producing a word-for-word translation, he or she should have professional work experience with translating or developing surveys.
  • A reviewer fluent in English and the target language to conduct an editorial and cultural language review of the translation. The reviewer should have strong language skills, knowledge of the culture of the survey population, and prior work experience in translation projects.
  • A designated decisionmaker to make final decisions about the pretest translation and the final survey translation when a consensus has not been reached by team members (i.e., an adjudicator). Only one decisionmaker is necessary if the person is fluent in both English and the target language and understands the intent of the survey items. There can be two decisionmakers if one has the necessary bilingual skills and the other the requisite understanding of survey item intent. Decisionmakers often include the project director as well as a university consultant or a person who helped to translate another AHRQ patient safety culture survey in the same target language.
  • Pretesters to conduct cognitive interviews or, if that is too costly or time consuming, to conduct a focus group to assess whether the translation meets its goals. Pretesters may include the bilingual reviewer or translator, both of whom are familiar with the survey, the translation goals, and specific translation issues that need to be explored during the pretest.

The size of your team can vary. For example, you may choose to have more than one translator or reviewer. As noted, some team members may fill more than one role. The translator and reviewer roles, however, should be filled by different persons.

Return to Contents

Step 2: Develop a Draft Translation

The translation coordinator collects, provides, and discusses the following materials and information with the translator, reviewer, and designated decisionmaker:

  • The most recent English-language version of the hospital, medical office, or nursing home AHRQ patient safety culture survey (available at http://www.ahrq.gov/qual/patientsafetyculture/).
  • An annotated dimensions document for the survey, which groups the survey items by the 12 dimensions they are intended to measure and gives additional information about the intended meaning of the items. We recommend using this document when translating the survey items because it is easier to understand the intended meaning of the items when they are grouped by dimension since they measure similar concepts. Translators can use the question number to determine the order in which the items appear in the formatted survey since they are NOT grouped by dimension in the formatted survey.
    • Example: Staff training in the nursing home survey refers to training provided by the nursing home, not formal training or certification required to meet initial hiring criteria.
  • Translators should refer to this document when they translate the survey.
  • If available, a copy of an existing reliable and validated translation of the AHRQ patient safety culture survey in your target language.
  • A clear statement of the translation goals—for example: to produce a translation that conveys the same meanings as the English version does, is easily understood by the target population, is culturally appropriate, and uses correct grammar and syntax.
  • Intended mode of survey administration (e.g., paper or Web, or both).
  • Sociodemographic characteristics of the target survey population (e.g., typical job titles, range of education levels, country or region of origin, age range; primary spoken/written language).
  • Target reading level (the reading level should be appropriate for most persons in the survey population).
  • Country, countries, or regions of a country where the survey will be administered.
  • Language/dialect to use (e.g., a standard-mix Spanish language that will be understood by speakers of Spanish from different countries versus a subgroup- or region- or country-specific language).

The translator, after discussions about the background materials:

  • Produces a translation that meets the translation goals, not a word-for-word translation that may not have the same meaning as the English version.
    • Example: In question B4 of the hospital survey B4 (My supervisor/manager overlooks patient safety problems that happen over and over), the translator needs to understand that the word "overlooks" means "ignores" or "does not pay attention to."
  • Documents translation issues and how they were addressed. The documentation can be done in various ways—for example, in a special table designed for that purpose or as comments inserted into the survey translation file. The table, or file, should contain both the English version and the translation.
  • Remains in contact with the translation coordinator during the translation process to address any issues that occur or to clarify the translation task.

Return to Contents

Step 3: Review Draft Translation

The bilingual reviewer:

  • Reviews the translation and the translator's documentation to assess whether the translation is accurate in meaning, grammar, and syntax and uses language that is familiar to and culturally appropriate for the target population:
    • Example: When talking about the nursing home survey item "Staff feel like they are part of team," it may be appropriate to translate the word "team" as "work team" if the concept of a "team" at work is not well established and respondents associate "team" with sports teams.
  • Indicates whether the translation of each item/instruction/response is acceptable or makes revisions. The reviewer also explains how the revisions will improve the translation—this documentation can be done in a table that includes the English version, the original translation, and any reviewer revisions and comments, or the reviewer can include comments in a tracked—changes file containing both the English wording and the translation.
  • Documents issues to address during the pretest:
    • Example: Explore what words Spanish-speaking employees typically use when speaking about all staff working in the nursing home—"el personal," "los empleados," or some other words?

Return to Contents

Step 4: Conduct Initial Review and Signoff

  • The team meets to review the translation and the reviewer's suggested changes and to make decisions about the pretest translation. If there is no team consensus about a translation issue, the decisionmaker reviews the translation and all documentation to decide on final changes to the pretest version of the translation.
  • The translation coordinator arranges for the approved draft to be prepared for use in a pretest.

Return to Contents

Step 5: Pretest the Translation

The next step is to pretest the draft translation with target population members (i.e., hospital, medical office, or nursing home staff). The pretest goal is to prevent response error by finding out if the translation is conceptually and culturally equivalent to the English version and easy for respondents to understand and answer. For the pretest, we recommend cognitive interviews with staff whose primary language is the target language. If time and resources are too limited to conduct cognitive interviews, consider conducting a focus group with the same types of staff. For information on conducting cognitive interviews, go to Appendix A. For information on conducting a focus group, go to Appendix B.

Return to Contents

Step 6: Conduct Final Review and Signoff

  • The team will have discussed the pretest findings (from either the cognitive interviews or the focus group) and reached consensus on a set of recommendations for revising the translation. The decisionmaker now makes final decisions on any issues lacking team agreement and signs off on revisions to the translation.
  • The translation coordinator arranges for the final version of the translation to be prepared and formatted for survey administration.

Return to Contents

Step 7 (Optional): Conduct Field Test and Psychometric Analyses

This recommended step is optional because of its cost and time requirements. If Step 7 is carried out successfully, however, the translation team will be able to assess the reliability and validity of its translation and its equivalence to the English version.

  • Conduct a field test of the survey with a selected group of hospitals/nursing homes/medical offices. The group can be a purposive sample but should still be broadly representative of major characteristics of the targeted health care setting.
  • Analyze response distributions and conduct psychometric analyses of the collected data.
  • Compare results with the results for the English version.

You are encouraged to E-mail a copy of your final translation and a description of the process used to develop it to:

SafetyCultureSurveys@ahrq.hhs.gov (Subject line: International translations)

Please indicate whether Westat may provide your contact information to other survey users interested in your translation.

Return to Contents

References

Agans RP, Deeb-Sossa N, Kalsbeek, WD. Mexican immigrants and the use of cognitive assessment techniques in questionnaire development. Hisp J Behav Sci 2006 28(2):209-30.

Comparative Survey Design and Implementation Initiative, Institute of Survey Research. Translation guidelines. In: Cross cultural survey guidelines: website user guide (Chapter VI). Ann Arbor: University of Michigan; 2008. Available at: http://ccsg.isr.umich.edu/translation.cfm.

Douglas SP, Craig, CS. Collaborative and iterative translation: An alternative approach to back translation. J Internat Mktg 2007 15(1):30-43.

Forsyth BH, Kudela MS, Levin K, et al. Methods for translating an English-language survey questionnaire on tobacco use into Mandarin, Cantonese, Korean, and Vietnamese. Field Methods 2007 19:264-83.

Krueger RA, Casey, MA. Focus groups: a practical guide for applied research, 4th ed. Thousand Oaks, CA: Sage; 2008.

Tourangeau R. Cognitive sciences and survey methods. In: Jabine T, Straf M, Tanu J, et al., eds. Cognitive aspects of survey methodology: building a bridge between disciplines. Washington, DC: National Academy Press; 1984. p. 73-100.

Census Bureau guideline: language translation of data collection instruments and supporting materials. Washington, DC: U.S. Census Bureau; 2009. Available at: http://www.census.gov/cac/2010_census_advisory_committee/language_translation_guidelines.html.

Weidmer B, Hurtado, M, Weech-Maldonado R, et al Guidelines for translating CAHPS® surveys. Rockville, MD: Agency for Healthcare Research and Quality; 2006.

Willis G. Cognitive interviewing: a tool for improving questionnaire design. Thousand Oaks, CA: Sage; 2005.

Return to Contents

Appendix A

Option 1 for Pretesting: How To Conduct Cognitive Interviewing

What is cognitive interviewing? Cognitive interviewing is a widely used method for pretesting and evaluating survey questionnaires that traditionally focuses on examining four stages of the question-answer process (Tourangeau, 1984):

  • Comprehension—how survey respondents interpret and comprehend the item.
  • Retrieval—how respondents search for information relevant to responding to the item.
  • Judgment—how respondents evaluate and integrate information retrieved to generate a response.
  • Responding—how respondents convert the internal response to one that is appropriate for the situation and understood by others.

For translations of the AHRQ patient safety culture surveys, interviewers should also focus on whether the translation:

  • Conveys the same meaning in the target language as in the English version.
  • Uses language that is culturally appropriate and familiar to the target population.

In-person cognitive interviews. For self-administered surveys, such as the AHRQ patient safety culture surveys, cognitive interviewing is usually conducted one-on-one during in-person interviews. Typical interviews are conducted as follows:

  • Interviewers train respondents to "think aloud" while they complete the survey. As respondents read the questions, they talk aloud about what they think the questions are asking and how they chose their answers.
  • Interviewers observe and listen.
  • At designated points in the survey (e.g., at the end of each survey section), the interviewers:
    • Ask open-ended questions to find out if anything in the survey is confusing, misinterpreted, or difficult to understand and answer.
    • Probe on specific words or phrases that may have a different meaning in the respondent's culture or be unfamiliar to the respondent.

Telephone cognitive interviews. Some of the cognitive interviews can be done by telephone if necessary to reach respondents in a wider geographic area. The translation is sent to respondents to complete prior to their interviews. They are asked to mark any words or items they found confusing or difficult to answer. If possible, respondents should return a copy of their completed surveys before their interviews for the interviewers to review. During the telephone interview, the interviewers probe about the survey questions and responses. As they do so, they need to refer to survey section and question numbers so that respondents can follow along in their copies of the survey.

How many interviews? Experience has shown that cognitive interviewing is most effective when results are analyzed and changes made during successive rounds of interviews (Willis, 2005, has a comprehensive discussion of cognitive interviewing).

You may need to go through your organization's Institutional Review Board process for approval to conduct this activity.

Cognitive interviewing involves the following activities:

Select Your Cognitive Interviewing Team and Method of Interviewing

  • Identify and recruit several available (preferably experienced) bilingual cognitive interviewers. Strongly consider using the translation reviewer as a cognitive interviewer because of that person's knowledge of the survey translation and the intent of the survey items.
  • Decide whether the interviews will be conducted in person or via the telephone, or both. If the interviews are conducted in person, arrange for appropriate space for conducting the interviews.

Develop Training Materials and Train Interviewers

  • Collect existing training materials, including:
    • General information about the survey.
    • A copy of the translation.
    • The annotated dimension version of the survey that focuses on the correct meaning of items.
  • Develop the following additional interview materials:
    • An Interviewer Guide, written in the target language, that contains the open-ended probes and more specific probes mentioned earlier:
      Examples: In your own words, what is this question asking? What words do you usually use to describe where you work when you are talking to friends and family?).
      Interviewers will also ask unscripted probes as needed during the interview.
    • A participant consent form in the target language. The form clearly but briefly states the purpose of the interview, the participant's role, and how the results will be used. We recommend that the form include a request to have the interview audio-recorded (to make sure that participants' comments are captured accurately). During telephone interviews, the form is read to participants and they are asked to give oral consent. After the recorder is turned on, they are asked again to give their consent to recording the interview.
    • Basic training materials on cognitive interviewing techniques and appropriate types of probes (a good resource is a how-to cognitive interviewing guide by Gordon Willis that is available at: http://appliedresearch.cancer.gov/areas/cognitive/interview.pdf (Plugin Software Help).
    • A report template that includes both the English version and the pretest translation version.
  • Conduct training on the interview materials. Include interviewer practice in administering the Interviewer Guide and using unscripted neutral probes.

Recruit Participants and Schedule Interviews

  • Recruit staff whose primary language is the target language. Try to recruit a mix of job titles and both men and women. Also, try to recruit staff who work different shifts (day, evening, night, weekends).
  • Allow sufficient time for recruiting—reaching staff and scheduling an interview can be difficult. Some interviews may need to occur before or after work hours, during lunchtime, or on weekends.
  • Send confirmation letters to recruited staff (also send the survey to telephone interview participants to complete before the interview). If the interviews occur more than a week after the confirmation letter is sent, remind the recruited staff 1 to 2 days before their scheduled interviews.
  • It may be necessary or appropriate to pay incentives to interview participants.

Conduct Interviews, Analyze Findings, and Recommend Changes

  • Conduct the interviews. Monitor early interviews to determine if any interviewer retraining is necessary.
  • Have the interviewers develop an interview summary for each of their interviews (the translation coordinator should provide a report template that includes both the English version and the pretest translation version).
  • Members of the translation team should analyze and discuss the findings after each round of interviews to determine whether to change the translation or the Interviewer Guide before conducting more interviews. You may want to designate a team member to analyze the interview summaries from a round and prepare a report for discussion with other team members. If you can conduct only one round of interviews, analyze early findings to make decisions about changes to the translation or Interviewer Guide before completing the round.
  • After all interview rounds have been completed, the team reviews all the findings, with emphasis on those from the last round of interviews. During the meeting, the team discusses key findings and translation issues and possible ways to resolve those issues.
  • Team members develop recommendations for changes to the translation, with explanations for the recommended changes. They note any translation issues they disagree about.

Continue with Step 6 of the translation process.

Return to Contents

Appendix B

Option 2 for Pretesting: How To Conduct a Focus Group

We recommend cognitive interviews for pretesting the translation. However, if you have limited time or resources, consider conducting a focus group (Agans, et al., 2006, discusses their experiences in conducting a focus group during the survey translation process). You may need to go through your organization's Institutional Review Board process for approval to conduct a focus group.

This pretest option involves the following activities.

Select Your Focus Group Team

  • Designate a focus group coordinator to handle logistics and oversee recruiting.
  • Select a moderator to conduct the focus group (can be a bilingual member of the translation team, such as the reviewer, or a professional moderator who is fluent in the target language). If a team member with little or no experience in conducting focus groups is selected, the person should review materials on conducting focus groups (e.g., Krueger and Casey [2008] or a similar source in the target language).
  • Assign a project team member to take notes during the focus group.

Prepare for the Focus Group

The focus group coordinator needs to:

  • Select a convenient time and facility. Typically, the facility will include a large conference table that seats 8 to 12 persons. You may need to offer child care services for group members with children.
  • Arrange for audio-recording equipment.
  • Serve refreshments to help participants relax and get comfortable talking with one another before the discussion begins.
  • Provide each participant with a name tent—you can ask participants to put either their first name or a pseudonym on the tent to make them more comfortable with speaking during the focus group.
  • Supervise and assist with preparation of materials for conducting the focus group.
    • If a professional moderator is selected who does not know about the project, discuss all survey background information and current task documentation with him or her, including translation issues and concerns.
    • Develop a Focus Group Guide in the target language for the moderator to use during the focus group. Questions should be designed to promote group interaction and a conversation specifically about the translation and possible problems in understanding, interpreting, or answering the questions. Also, questions should be asked to assess the cultural equivalence of the translation to the AHRQ version of the patient safety culture survey.
    • Develop a participant consent form in the target language. The form clearly but briefly states the purpose of the interview, the participant's role, and how the results will be used. The form should also ask for permission to audio-record the focus group session. Participants give their consent by reading and signing the form.
  • Recruit staff (hospital, medical office, nursing home staff) whose primary language is the target language. Inform staff that the focus group session will be audio-recorded to capture an accurate record of participants' comments. Try to recruit staff with a mix of job titles and a mix of shift schedules. Because participants often cannot participate after agreeing to do so, recruit two to three more participants than you need. Incentives are usually paid to focus group participants at the session.
  • Schedule the focus group, send confirmation letters to recruited staff, and remind the recruited staff 1 to 2 days before the focus group.

Conduct the Focus Group, Analyze the Findings, and Develop Recommendations for Changes

  • The moderator conducts the focus group, and the notetaker records participants' comments.
  • The focus group coordinator arranges for a transcript of the audio-recording.
  • Team members analyze participant comments and document key findings, such as participant problems understanding questionnaire items, words, instructions, or response options.
  • Team members develop recommendations for changes to the translation, with explanations for the recommended changes. They note any translation issues they disagree about.

Continue with Step 6 of the translation process.

Current as of March 2010


Internet Citation:

Translation Guidelines for the Surveys on Patient Safety Culture. March 2010. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/qual/patientsafetyculture/transguide.htm


 

Current as of March 2010
Internet Citation: Translation Guidelines for the Surveys on Patient Safety Culture. March 2010. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/professionals/quality-patient-safety/patientsafetyculture/transguide.html