Abstract
PURPOSE We assessed whether existing guidelines for the primary care of lesbian, gay, and bisexual (LGB) people meet appropriate standards of developmental rigor, and whether they provide consistent recommendations useful for primary care clinicians.
METHODS We performed a systematic review of such guidelines using the Cochrane Collaboration method. The countries searched were Australia, Canada, Ireland, New Zealand, the United Kingdom, and the United States. For sources, we used electronic databases, guidelines databases, primary care professional organizations, government departments of public health, LGB health care textbooks, and national LGB organizations. We assessed the quality of existing guidelines using the validated Appraisal of Guidelines for Research and Evaluation (AGREE) instrument and compared the recommendations from all fully appraised guidelines.
RESULTS Our search did not identify any previous systematic reviews on primary care of LGB people. Of 2,421 documents identified, we initially reviewed 30 and fully appraised 11, none of which completely satisfied the AGREE criteria for quality and only 2 of which were specifically designed for primary care. Developmental rigor was poor. Particular gaps were a lack of explicit inclusion criteria, independent reviewers, and updating procedures. Nonetheless, we did identify several consistent recommendations pertinent to primary care settings: guidance on inclusive clinical environments, standards for clinician-patient communication, sensitive documentation of sexual orientation, knowledge for cultural awareness, staff training, and addressing population health issues.
CONCLUSIONS Currently available guidelines for LGB care are philosophically and practically consistent, and provide a degree of evidence-based clinical and systems support to primary care clinicians. There is a need, however, for evidence-based LGB guidelines that are more rigorously developed, disseminated, and evaluated specifically for the primary care setting.
- Homosexuality, male
- homosexuality, female
- bisexuality
- female
- male
- minority groups
- delivery of health care
- standards
- guidelines
- evidence-based medicine
- primary care
INTRODUCTION
The primary care of lesbian, gay, and bisexual (LGB) people is compromised by gaps in clinical care and practice systems.1 These gaps include documented deficiencies in the LGB-specific knowledge and skills of health care professionals,1 which are, in part, explained by a lack of training about LGB health at all levels of their education.2,3 Clinician knowledge is also limited by the dearth of available population-based data,4 although such studies are increasingly being conducted.5 Further, the practice environment is affected by the contentious and stigmatized nature of homosexuality, with health care professionals holding a range of beliefs about minority sexual orientation that are occasionally pathologizing and commonly minimizing.6 Clinician beliefs that sexual orientation is irrelevant form a major barrier to patient disclosure within consultations,7 and few primary care services create a practice environment that demonstrates awareness of and respect for LGB patients as a group.8 This lack of awareness and respect has been described as a “blind spot” among family physicians toward LGB patients,9 and it leads to reports by LGB people of difficulty in accessing culturally competent primary care services.10
LGB-specific services have been developed to overcome these deficiencies11; however, there is a pressing need for mainstream services to become culturally competent in this area.12,13 We assert that such cultural competence should be LGB-specific rather than generic to overcome a tendency to disregard minority sexual orientation in clinical care. Many family physicians prefer to avoid stereotyping and remain neutral by ignoring numerous specific cultural attributes; as a result, crucial differences that influence health are missed.14
Clinical practice guidelines for LGB health care would assist in improving clinicians’ knowledge and skills15; however, few guidelines are available, creating a further barrier to effective practice.2 Policies on lesbian and gay health do exist within some medical associations,16,17 although the associations have been slow to translate these policies into standards of practice and education.18 Health policy at the government level also often lacks LGB-specific focus, which further discourages LGB sensitivity among educators or clinicians.19 To be most effective, guidelines must be accompanied by a range of other educational methods including audits, training, and feedback20; nevertheless, they can serve to raise the awareness of clinicians to the LGB population group. Guidelines that provide easy access to current evidence in the area and clear recommendations on day-to-day care are more likely to be put into practice by clinicians.21,22
We undertook a systematic review to determine whether existing LGB guidelines for primary care meet appropriate standards of rigor in their development and to summarize areas of consistency that provide practical guidance for primary care clinicians.
METHODS
Our review process was based on that outlined in the Cochrane Handbook for Systematic Reviews of Interventions.23
Search Strategy
We limited our search to 6 source countries—Australia, Canada, Ireland, New Zealand, the United Kingdom, and the United States—as these are the major English-speaking countries with primary care systems that have some similarity. We used multiple data sources and search terms as shown in Table 1⇓. The search terms were designed to systematically identify documents with a focus on LGB people in primary care settings. We did not focus on transgender or intersex individuals, although some of the documents included these populations. After finding few relevant guidelines through the peer-reviewed data sources, we next searched Web sites of general practice and nursing primary care organizations in each country, and health department Web sites. We also identified relevant textbooks published since 1995 through Google Book.
Inclusion Criteria for the Initial Review
We used the validated Appraisal of Guidelines for Research and Evaluation (AGREE) instrument, which has been designed to assess guidelines for the quality of reporting, the quality of the recommendations, and the likelihood that they will achieve the intended outcome.24 The instrument has been increasingly used by guideline developers and collators, including the Guidelines International Network.25 It contains 23 items in 6 domains, and inclusion criteria for our initial review related to 4 of these domains: Scope and Purpose, Stakeholder Involvement, Rigor of Development, and Clarity and Presentation. Our specific inclusion criteria were an explicit LGB and primary care scope and purpose; at least some involvement of stakeholders, including health care professionals, researchers, and LGB consumers; developmental rigor that demonstrated explicit links between research evidence about LGB health and the recommendations; and clear presentation pertaining to the practice environment and specific knowledge for LGB cultural awareness. We did not regard the final 2 AGREE domains of Applicability and Editorial Independence as inclusion criteria for initial review as they were very rarely addressed.
Selection Process for Full Appraisal
One of the authors (R.P.M.) reviewed all abstracts of the papers found on the initial search and excluded those that did not satisfy the LGB and primary care scope. Both authors then independently read the full documents identified for an initial review and excluded those that did not meet our inclusion criteria in the 4 AGREE domains. Several documents were excluded on the basis of poor methodology, inadequate or nonexistent involvement of stakeholders, limited or no specific recommendations, or some combination thereof.
Both authors fully appraised the remaining documents independently, which followed the AGREE recommendation that each of the reviewed guidelines be assessed independently by at least 2 reviewers. Each reviewer independently rated the document on the items in the 6 domains using a 4-point scale: 4 = strongly agree, 3 = agree, 2 = disagree, 1 = strongly disagree. We then calculated a score for each domain, and met to compare scores and calculate the standardized score of each domain for each document. The formulas for calculating the scores for each domain are as follows: the maximum possible score was 4 × the number of items × the number of appraisers; the minimum possible score was 1 × the number of items × the number of appraisers. Using these values, the standardised domain score expressed as a percentage is the obtained score minus the minimum possible score, divided by the maximum possible score minus the minimum possible score. Hence, the final domain score was the mean percentage score of the 2 reviewers according to the AGREE scoring system, with possible scores ranging from 0% to 100%.24 The higher the percentage, the higher the quality of the guideline for that AGREE domain.
Finally, we considered whether a distribution plan was in place, as there is evidence that guidelines are more influential if they have been endorsed by professional bodies and undergone widespread distribution and piloting.15,22
RESULTS
Review Outcome
We identified 2,421 publications on the initial search, 30 of which fulfilled our minimum criteria for initial review, and 11 of which satisfied our inclusion criteria for full appraisal (Table 1⇑). Five sources generated the 11 documents. One document was from a primary care organization, the Royal College of Nursing in the United Kingdom.26 Four documents were from government departments of health in Massachusetts,27 Washington State,28 Scotland,29 and British Columbia.30 A US private health care agency, Kaiser Permanente, had also produced a clinician’s handbook on LGB and transgender culturally competent care.31 Two textbooks fulfilled the criteria for full appraisal,2,32 as did 3 sets of documents from leading LGB organizations in Australia,33–35 the United States,36 and Ireland.37 We have summarized the characteristics of the 11 fully appraised documents in Table 2⇓.
Guideline Quality According to Appraisal Scores
We obtained a high level of agreement on our independent AGREE domain scores for each fully appraised document, with all but 2 scores being within 2 points. We then calculated the combined standardized AGREE scores as a percentage for each document, as recommended by the AGREE developers (Table 3⇓). None of the selected guidelines fully satisfied the AGREE criteria in all domains. The highest standards were achieved for the Scope and Purpose domain (indicating that guidelines were about LGB people and could apply to primary care), which was expected as this was our primary inclusion criterion for initial review. Only 1 textbook32 and the Irish document37 were specifically designed for primary care.
Scores for the Stakeholder Involvement domain varied widely. Most documents scored well on clearly defining the target users. Individuals from relevant professional groups who were involved in the development were usually just a small range of expert clinicians, researchers, or both. Just 3 documents sought patients’ views and preferences.27,29,34 The best examples in this domain were the Justice Resource Institute engagement of a multidisciplinary working group of more than 60 people, including consumers,27 and a wide-ranging consultation process with consumers in Scotland.29 Only 1 group29 pilot tested its guide among target users.
The clarity of presentation quality also varied and was best in the 2 textbooks,2,32 with their large format; extensive resource lists, tables, and examples of forms; and inclusion of useful summaries at the end of each chapter. Three of the documents2,28,35 were accompanied by extensive resources on a linked Web site, which added to their presentation quality and availability. The final domains of Applicability and Editorial Independence were poorly addressed in all of the documents. Low applicability, in particular, creates a practical problem in cases where services and clinicians might want to use the guidelines, but the guidelines do not list barriers, costs, or review criteria.
All of the fully appraised documents scored poorly in the Rigor of Development domain as well. We describe the documents’ performance on each of the 7 items in this domain below, as we consider the evidence base of guidelines to be crucial to their value in the clinical setting given the difficulty clinicians have in accessing evidence regarding LGB health.
-
First, none of the guidelines documented use of systematic methods for their literature review. As a result, guidelines could be subject to bias or reflect a particular assumption of the authors. For example, the Irish guidelines contained a statement that “being lesbian, gay or bisexual is not a preference and is not chosen. It is integral to a person’s…identity rather than being a lifestyle.”37 This statement reflects just 1 essentialist perspective without offering an alternative social constructivist perspective that has equal validity in the literature.38 Further, there was a tendency for North American documents to restrict their literature reviews to local studies and exclude relevant international work.
-
Second, the Rigor domain requires that the inclusion and exclusion criteria for selecting evidence be described, which was not done in any document. Two documents referred to levels of evidence and occasionally described them, but did not have explicit criteria.34,37
-
Third, a description of how the recommendations were formulated was included in only 4 documents, and these descriptions were not detailed.27,29,32,34 The methods used were expert opinion,32 consensus following empirical evidence from interviews or focus groups,27 or consensus following stakeholder consultation.29,33
-
Fourth, all but 3 documents mentioned the benefits of their recommendations; however, none considered risks or alternative possibilities.
-
Fifth, the recommendations should be explicitly linked with the evidence. Scores within the domain were highest for this item, as 5 documents provided excellent or very good links.2,27,31,32,37 But 4 documents did not provide any links at all.
-
Sixth, the guidelines should be reviewed by an independent expert before publication, but none of the documents stated that such review had been done.
-
Finally, the guidelines should include a procedure for updating, including a time line, and just 1 document complied in this regard.29
In addition to assessing the AGREE criteria, we assessed whether the documents contained any process for dissemination to or engagement of primary care clinicians. Three of the documents had a dissemination strategy. Kaiser Permanente distributed its standards to all of its own health care professionals as well as to all medical schools in the United States.39 The UK Royal College of Nursing standards gained endorsement from 4 allied health colleges, and from the sponsoring union (UNISON), which assisted in distribution.26 The Irish document37 had the most comprehensive process. This document was launched at the Irish College of General Practitioners’ annual conference and then sent to every general practitioner in Ireland by the College (e-mail communication, Odhrán Allen; March 9, 2010). Moreover, some of the guidelines have been incorporated into the Irish general practitioner training curriculum.
Synthesis of Recommendations
We compared the recommendations within the 11 fully appraised documents and conducted a thematic and content analysis. All documents used common principles of cultural competency, equity, or the need for awareness of LGB patients and their diversity. We also found considerable concordance among the documents on many of the individual recommendations. The majority of documents contained recommendations that related to both practice systems and clinical care. We noted several consistent themes regarding practice systems: inclusive clinical environments; documentation; staff and training; and population health. Themes pertaining to clinical care were areas of LGB-specific knowledge, and clinician-patient communication. These common themes and recommendations are shown in Table 4⇓.
An important weakness was the limited or lacking guidance provided for clinicians on disclosure of sexual orientation, which has been found to be a considerable challenge for patients and clinicians alike.7 Very few documents focused on how to facilitate disclosure, and only 3 discussed the need to respect the patient’s choice not to disclose (Table 4⇑).28,31,36 In addition, few included recommendations on documentation issues, and none mentioned LGB sensitivity in referral letters. Only 4 documents recommended forming a partnership with LGB agencies, despite the importance of such partnership for the provision of resources and for referral to LGB community services.27,29,30,34 Finally, a population health focus was unusual, and in particular, only 3 documents noted the need for clinicians to advocate on behalf of LGB patients.27,31,32
DISCUSSION
This is the first systematic review of LGB guidelines that we are aware of. Our review was strengthened by the broad range of data sources accessed, including government and LGB organization Web sites, which turned out to be the sources of the majority of the fully appraised documents. A further strength was the use of the validated AGREE instrument to appraise documents and so provide an indication of their quality. In addition, the rigor of the review process was enhanced by having 2 reviewers conducting independent appraisal. Although the broad range of data sources was a strength, it arose from a limitation. Our initial search of databases revealed a complete absence of LGB-focused guidelines, which necessitated a search of individual organizations and government departments. This lack of guidelines in commonly used databases highlights the difficulty that health care professionals and organizations are likely to have in finding relevant guidelines. It also led to a limitation of this review in that inevitably, we were unable to find all relevant guidelines because we could not search every government department or LGB agency in the countries of interest.
We synthesized the recommendations into a cohesive and, we suggest, clinically useful set of guidelines for the primary care of LGB people. Such guidelines have the capacity to be “a potential change agent in the health care system”20 by enabling clinicians and managers to translate current research evidence into clinical practice.20 Conversely, guidelines that contradict existing values of clinicians, or highlight new knowledge and skills that are needed can “demand too much change to existing routines” 22 and therefore reduce the likelihood of guidelines uptake by clinicians.22
We suggest that LGB guidelines development should be improved in the following ways to ensure reliability and uptake. Involvement of LGB stakeholders in consultation would assist in addressing diversity within LGB groups as well as ensuring that issues of clinical environment, confidentiality, and communication are truly patient centered. Rigor would be improved by the use of a systematic review, which would ensure inclusion of international LGB research and address multiple theoretical perspectives. Another essential element in presentation of the recommendations is to reference all recommendations with the evidence on which they are based.
A further method to improve the rigor of guideline development is to detail the inclusion and exclusion criteria for the review, particularly as minority sexual orientation is such a multidimensional phenomenon. For example, guidelines should state whether transgendered gay and lesbian people, or bisexual women, or heterosexually identified men who have sex with men are included as a target group. External review by an independent expert would also improve the perceived reliability of the guideline. We suggest that the most effective reviewers for primary care LGB guidelines would be mainstream researchers or practitioners in the primary care field. The Irish guide came closest to this model in that a gay and lesbian organization researched and wrote the document, which was then reviewed by the leading body for general practice.37 Finally, the field of LGB health research is expanding rapidly, so it would be reasonable to include a recommended time line of about 5 years before the guideline is updated with new evidence.
Implications for Clinical Practice and Future Research
The synthesized recommendations from the appraised documents could be readily applied to existing primary care systems. Most of the recommendations would require minimal change and no cost to practice systems, which are prerequisites for uptake of guidelines by primary care clinicians.22 The areas of special knowledge for LGB health care contained in the recommendations would improve the LGB cultural competence of clinicians that LGB clients expect. For example, clinicians could make minor changes to their communication style and language that would help LGB people to feel more comfortable. The recommendations could also provide primary care services with ways to improve policies and procedures and local LGB population health initiatives. At the policy level, they could encourage the development of guidelines targeted to specific areas of the health system including primary care.
The gaps identified in the existing guidelines could be addressed within future research, in particular focusing on the primary care setting. For example, most of the guidelines reviewed made general statements pertaining to the LGB group as a whole, rather than highlighting specific health care needs according to sex, diverse expression of sexual orientation, socioeconomic status, age, and ethnicity. Further, research that examines the complexities of disclosure of sexual orientation would add much needed depth to future guidelines.
Future LGB guidelines should be specifically and rigorously developed for primary care. Primary care organizations should be involved in this process to ensure that the guidelines satisfy their local needs and are disseminated widely. Finally, an evaluation and review process must be built into the implementation plan to understand whether the desired improvement in access, cultural competence, and quality health care has been achieved.
Footnotes
-
Conflicts of interest: none reported
-
Funding support: The first author received a National Health and Medical Research Council (NHMRC, Australia) Primary Care Research PhD Scholarship, and a research grant from the Royal Australian College of General Practitioners, which supported the conduct of this systematic review.
- Received for publication August 28, 2009.
- Revision received June 11, 2010.
- Accepted for publication June 23, 2010.
- © 2010 Annals of Family Medicine, Inc.