Barriers to physician adherence to a subfertility guideline

Abstract

BACKGROUND: Guidelines aim to improve clinical practice but are not self-implementing. Insight into barriers to physician guideline adherence is crucial for development of effective implementation strategies. The study aim was to identify barriers to physician adherence to an intrauterine insemination (IUI) guideline of the Dutch Society of Obstetrics and Gynaecology. METHODS: We conducted a cross-sectional survey among all Dutch gynaecologists, residents and fertility physicians (n = 860), using written questionnaires that were based on information obtained in focus group discussions. We investigated barriers related to physicians’ knowledge and attitudes, and external barriers. RESULTS: The response rate was 65%. We used 344 questionnaires for analysis. Physicians’ knowledge was adequate, with a median unfamiliarity rate with each of the 31 key guideline recommendations of 12%. Physicians’ attitudes were generally positive, especially regarding guideline development and quality. Important attitude-related barriers included physicians’ lack of self-efficacy regarding physician–patient communication and poor outcome expectancy. External barriers were mostly related to specific patient characteristics and associated with higher age of physicians and fewer consultations for fertility problems per week. CONCLUSIONS: Multiple barriers impede physician adherence to subfertility guidelines, mainly physicians’ lack of self-efficacy and low outcome expectancy. Both physicians and patients play an important role in future implementation interventions to optimize subfertility care.

Introduction

In the field of reproductive medicine, advanced diagnostic and treatment techniques are associated with high physical and psychosocial impacts on patients; risks of various complications, such as ovarian hyperstimulation syndrome and multiple pregnancies; ethical dilemmas; and high costs (Hughes, 2003; Pandian et al., 2003). Subfertility guidelines, such as the fertility guideline of the National Institute for Clinical Excellence (NICE), aim to assist physicians and patients in the decision-making process regarding appropriate, safe and cost-effective care and to improve the quality of subfertility care (Field and Lohr, 1990; Woolf et al., 1999).

However, guidelines are not self-implementing (Grol, 1997). Guideline implementation may be best described as ‘a stepwise and planned introduction of a guideline, aiming to integrate its recommendations profoundly into the daily performance of health-care professionals’ (Grol and Wensing, 2001). Studies suggest that, besides publication and dissemination of guidelines, more intensive intervention strategies are necessary to promote guideline implementation and alter health-care delivery (Oxman et al., 1995; Bero et al., 1998; Grimshaw et al., 2004). It is increasingly recognized that these intervention strategies should be based upon assessment of potential barriers to guideline adoption (Grol, 1997; Grol and Grimshaw, 2003; Grol and Wensing, 2004).

A wide range of factors has been identified as possible barriers to the implementation of guidelines. A review showed that many investigators have focused on characteristics of individual physicians to explain failure of guideline implementation (Davis and Taylor-Vaisey, 1997). Cabana and colleagues reviewed 76 studies on barriers to physician guideline adherence and identified seven general types of barriers (Cabana et al., 1999). They developed a framework in which these barriers are classified into three main categories: barriers related to physicians’ knowledge (i.e. lack of awareness and lack of familiarity); barriers that affect physicians’ attitudes (i.e. lack of agreement, lack of self-efficacy, lack of outcome expectancy and lack of motivation); and external barriers (i.e. patient-, guideline- and environment-related factors).

Since most subfertility guidelines have only been disseminated and have not been systematically implemented yet, physician adherence to subfertility guidelines is probably not ideal. However, little is known about the extent of physician adherence to subfertility guidelines and the potential barriers involved. For instance, we still lack this information for the nine subfertility guidelines produced as part of the guideline programme of the Dutch Society of Obstetrics and Gynaecology (NVOG). The NVOG has issued 55 clinical practice guidelines since 1996, using a systematic approach to prepare evidence-based guidelines of high quality. The Appraisal of Guidelines for Research and Evaluation (AGREE) instrument (The AGREE Collaboration, 2003) is used to support this developmental process. NVOG guidelines are disseminated among NVOG members by mail and the internet (http://www.nvog.nl).

In this study, we focused on the NVOG guideline regarding intrauterine insemination (IUI). This guideline includes 31 key recommendations about indications, management and organizational aspects of IUI care (Table I). The main issues addressed in the IUI guideline are also within the scope of other subfertility guidelines, such as the NICE guideline. We aimed to identify barriers to physician adherence to this subfertility guideline and to classify the identified barriers according to the framework of Cabana and colleagues (Cabana et al., 1999). Additionally, we explored characteristics of physicians and practice settings associated with barriers to adherence to this subfertility guideline.

Materials and methods

Study design and population

We conducted a cross-sectional survey using written questionnaires among all 860 physicians working in the field of obstetrics and gynaecology in the Netherlands, comprising 698 gynaecologists, 82 residents and 80 fertility physicians (i.e. medical doctors specialized in assisted reproduction). Databases of the NVOG and the Dutch Society of Fertility Physicians (VVF) were used to obtain the contact details of the study population.

After the development and pilot testing of the questionnaire, we sent this questionnaire with an accompanying informative letter and a return envelope to all selected physicians. Two weeks after the initial mailing, we sent a reminder card to non-responders requesting them to complete and return the questionnaire. Subjects received a second copy of the questionnaire if they did not return the first questionnaire within four weeks.

Questionnaire development

A systematic review of the literature revealed that little is known about barriers to physician adherence to subfertility guidelines. Therefore, we used qualitative research methods in the developmental process of the questionnaire to obtain information about potential barriers to physician adherence to the IUI guideline. We performed two semi-structured focus group discussions, one with 13 residents and fertility physicians and one with 11 gynaecologists. These physicians were all commonly participating in IUI programmes. We selected participants of different age, sex, length of time in practice and practice setting in the south-east of the Netherlands. An independent gynaecologist used a topic guide with open-ended questions to moderate the focus group discussions. Questions were based on the barrier classification system described by Cabana and colleagues (Cabana et al., 1999). Barriers to physician guideline adherence were discussed with regard to the 31 key recommendations of the IUI guideline. Each focus group discussion lasted for approximately 2½ hours. Sessions were tape-recorded and fully transcribed. Two researchers (E.H., W.N.) conducted independently a content analysis. Potential barriers to physician adherence to the IUI guideline were identified and classified into the three previously mentioned domains of the framework of Cabana and colleagues (Cabana et al., 1999). Differences in interpretation were minimal and consensus was promptly achieved. For the development of our questionnaire, data from the focus group discussions were combined with an existing validated questionnaire designed to identify barriers to physician adherence to clinical practice guidelines (Peters et al., 2003).

Questionnaire content

The 19-page questionnaire consisted of 95 questions. To evaluate physicians’ knowledge, respondents were questioned about their familiarity with each of the 31 key recommendations of the IUI guideline by closed questions (yes/no). To assess physicians’ attitudes, we asked participants to respond to statements about agreement with each of the 31 key recommendations of the IUI guideline; agreement with the IUI guideline; self-efficacy (i.e. confidence in the ability to perform a behaviour); outcome expectancy; and motivation to follow guideline recommendations. To identify external barriers, we questioned physicians about their opinions regarding patient-, guideline- and environment-related factors affecting their behaviour to adhere to guideline recommendations. Opinions regarding attitude-related and external barriers were scored on a five-point Likert-scale (1 = strongly agree; 5 = strongly disagree). Suggestions for additional guideline recommendations and the prevalence of demographic characteristics were identified with closed questions with several answer possibilities.

Analysis

We entered data from collected questionnaires into a database. Data were analysed with the Statistical Program for the Social Sciences (SPSS 11.0 for Windows^®, SPSS Inc., Chicago, Illinois, USA). Frequencies were calculated to describe the main barriers to physician adherence to the IUI guideline.

We aimed to assess the independent effects of different determinants on barriers to physician guideline adherence, i.e. characteristics of physicians and practice settings, such as age, sex, education, speciality (gynaecologist, resident or fertility physician), type of gynaecologist (specializing or not in the field of subfertility care), average number of consultations for fertility problems per week, teaching hospital and IVF facilities, using multivariate linear regression analysis. First, we performed a factor analysis to cluster the different attitude-related and external barriers in order to reduce the number of dependent variables. Attitude-related barriers regarding agreement with each of the 31 key recommendations of the IUI guideline were clustered into six factors: indications (two factors); medication; cancel criteria; timing using LH measurements; and organizational aspects. Other attitude-related barriers (listed in Table II) were aggregated into three factors: agreement with the IUI guideline and outcome expectancy; self-efficacy; and motivation. External barriers (presented in Table III) were also clustered into three factors: patient-related external barriers (two factors); and guideline-related and environment-related external barriers. Subsequently, we calculated sum scores for all 12 factors of the factor analysis to obtain continuous dependent variables. Finally, Spearman’s correlation analysis was carried out to evaluate collinearity between determinants. All physician and practice setting characteristics mentioned above were selected for multivariate linear regression analysis with backward elimination procedures, except education and type of gynaecologist, owing to high correlation with speciality. Statistical significance was set at P < 0.05.

Results

Study population

Of the 860 distributed questionnaires, 11 were undeliverable. A total of 554 (65%) of the 849 physicians completed and returned their questionnaire. Of these, we excluded 210 physicians who reported that they never counselled subfertile couples about IUI and never performed IUI treatments. As a result, 344 questionnaires were available for analysis. Overall, 257 (75%) of the 344 participating physicians were gynaecologists, 40 (12%) residents and 47 (14%) fertility physicians. Among the gynaecologists, 85 (33%) specialized in the field of subfertility care. The demographic characteristics of the participating physicians are presented in Table IV.

Knowledge-related barriers

Physicians’ knowledge of the IUI guideline was adequate. Just 6% of the 344 physicians indicated that they lacked sufficient knowledge to perform according to the IUI guideline. Furthermore, the median level of unfamiliarity with each of the 31 key recommendations was 12% (range 2–49%). Only two guideline recommendations were exceptionally unfamiliar. Nearly half of the physicians (49%) were unaware that patients undergoing unstimulated IUI should be monitored using LH measurements twice daily. The other unfamiliar recommendation (41%) concerned the prescription of clomiphene citrate for controlled ovarian hyperstimulation in IUI treatment at a daily dose of 100 mg.

Attitude-related barriers

Physicians’ attitudes about the content of the IUI guideline were generally positive. The median level of disagreement with each of the 31 key recommendations was 7% (range 0–31%). We found that the two previously mentioned unfamiliar recommendations had high rates of disagreement among physicians: 31% mentioned lack of agreement with the recommendation to use LH measurements twice daily to monitor patients undergoing unstimulated IUI, and 26% disagreed with prescribing clomiphene citrate at a dose of 100 mg daily for controlled ovarian hyperstimulation in IUI treatment. Another key recommendation lacking agreement was related to the start of IUI treatment in couples with unexplained subfertility. Thirty-one per cent of the physicians disagreed with the guideline to start IUI treatment after 3 years of unexplained subfertility if female age is less than 36 years.

Statements of physicians about attitude-related barriers for IUI guideline adherence associated with agreement with the IUI guideline, self-efficacy, outcome expectancy and motivation to follow guideline recommendations are listed in Table II. First, lack of agreement with the IUI guideline was limited. Just 0.3% of the 344 physicians felt that this subfertility guideline was not developed rigorously. No more than 3% reported that the IUI guideline is not sufficiently evidence-based. In addition, a minority (3%) indicated that the IUI guideline does not contribute to uniformity in IUI care.

However, disagreement with certain aspects of the IUI guideline was also reported. The majority of physicians (57%) mentioned disagreement in relation to lack of reimbursement and pointed out that working according to the IUI guideline requires financial compensation. Eighteen per cent of the physicians believed that this subfertility guideline could be abused easily in medical disciplinary law. Physicians also reported limitations with regard to their flexibility in considering patients’ requests (19%) and making personal considerations (12%). Moreover, 14% mentioned that the IUI guideline does not consider characteristics of individual patients adequately. Indeed, 86% of the physicians suggested at least one additional recommendation to the IUI guideline and most suggested recommendations were related to patients’ characteristics; for example, body mass index (60%), early follicular phase FSH levels (47%), female age (37%), successful IUI treatment in the past (22%) and endometriosis (18%).

Low self-efficacy could also affect guideline adherence. Several physicians described lack of confidence in their ability to convince patients about both low success rates (24%) and risks of multiple pregnancies in IUI treatment (20%). Furthermore, 18% of the physicians felt unable to deny patients’ requests.

We found poor outcome expectancy among 29% of the physicians. These physicians mentioned that working according to the IUI guideline will not lead to improved outcomes in the short term.

Although only a minority (3%) reported resistance to working according to guidelines in general, 11% of the physicians mentioned poor motivation to perform according to the IUI guideline, because this would be time-consuming. Additionally, 9% of the physicians indicated difficulty changing well-established routines as a motivation-related barrier to applying this subfertility guideline.

External barriers

External barriers may affect the ability of physicians to perform according to the IUI guideline (Table III). Questioning physicians about patient-related external barriers revealed several specific characteristics of patients that complicate IUI guideline adherence, such as elevated early follicular phase FSH levels (43%), female age above 35 years (33%) and unexplained subfertility (22%). Important environment-related external barriers mentioned by physicians include the cost price of urine-based LH tests (29%) and lack of facilities in the practice setting (17%).

Determinants of barriers to physician adherence to the IUI guideline

In multivariate linear regression analysis, several physician and practice setting characteristics were found to contribute independently to seven out of 12 factors of clustered attitude-related and external barriers. However, for most determinants of these barriers the amount of variance explained in the final regression model was low. Determinants of barriers with more than 5% of the variance explained are presented in Table V. Fewer consultations for fertility problems per week proved to be a determinant of attitude-related barriers regarding motivation. This determinant is also associated with patient-related external barriers. Furthermore, we observed that working in a non-teaching hospital is related to attitude-related barriers regarding motivation. Higher age of physicians is a determinant of patient-related external barriers.

Discussion

This is the first study to identify barriers to physician adherence to a subfertility guideline and to explore determinants associated with these barriers. This survey revealed that a barrier with an important impact on adherence to subfertility guidelines is physicians’ lack of self-efficacy regarding communication with patients during the decision-making process about appropriate subfertility care. Physicians reported lack of confidence in their ability to convince patients about low success rates and risks of multiple pregnancies in IUI treatment and to deny patients’ requests. Physicians’ poor confidence in their convincing power may be attributable to inadequate knowledge of guideline recommendations, lack of communication skills or the persistence of external barriers (Cabana et al., 1999; Grol and Wensing, 2001; Cabana and Kim, 2003).

First, physicians may be unfamiliar with the specific content of a subfertility guideline (Cabana and Kim, 2003). A review of 31 surveys on physicians’ familiarity with guideline recommendations found a median unfamiliarity rate of 57% (Cabana et al., 1999). However, in our study the median level of unfamiliarity was only 12%. Although this unfamiliarity rate is based on physician self-report, which might be inaccurate, others demonstrated that physicians’ self-reported familiarity adequately reflects actual familiarity with guideline recommendations (Cabana et al., 2001). As a result, potential gaps in physicians’ knowledge about details of the IUI guideline are unlikely to contribute significantly to physicians’ lack of confidence in their convincing power.

Second, physicians involved in subfertility care may lack good communication skills (Grol and Wensing, 2001). Despite the fact that only 2% of the physicians in our study indicated that they lacked skills to carry out IUI guideline recommendations, it is not imaginary that more physicians have insufficient communication techniques, because judgements regarding personal skills and performance are generally not very accurate (Allen and Rashid, 1998). Education and training programmes that focus on the development of communication skills of physicians may be useful in improving physicians’ low confidence in their convincing power due to insufficient communication techniques (Hulsman et al., 1999).

Third, the persistence of patient-, guideline- or environment-related external barriers may eventually affect physicians’ confidence in their convincing power (Cabana et al., 1999). Patient-related external barriers may play a role if patients are opposed to or perceive no need for guideline recommendations. The inability of physicians to reconcile preferences of these patients with guideline recommendations may in time compromise their confidence in their ability to communicate with patients with conflicting ideas (Cabana et al., 1999). Interestingly, in our survey physicians mentioned several patient-related external barriers, especially regarding patients with specific physical and demographic features, such as elevated early follicular phase FSH levels, female age above 35 years or under 25 years, unexplained subfertility, and non-Dutch ethnic background. In particular, older physicians and physicians with fewer consultations for fertility problems per week reported these external barriers. Additional studies of patients’ views about subfertility care are imperative to elucidate possible discrepancies between patients’ preferences and subfertility guideline recommendations. Subsequently, patients’ preferences about subfertility care that interfere with recommendations of subfertility guidelines could be addressed in a patient-directed intervention. For instance, targeted education programmes for patients could improve their understanding of subfertility care and tackle possible misconceptions and unrealistic expectations. Additionally, an intervention could also focus on educating physicians about patients’ views about subfertility care and shared decision-making. Both interventions could facilitate the medical decision-making process and improve physicians’ confidence in their convincing power.

No major guideline-related external barriers were identified in our study. However, the IUI guideline does not contain recommendations about information delivery and counselling; this contrasts with the NICE guideline, in which the importance of shared decision-making is clearly acknowledged. Adding such recommendations to the Dutch IUI guideline, especially recommendations regarding the counselling of patients with preferences about subfertility care that impede guideline adherence, could be valuable for physicians and effective in resolving lack of self-efficacy as a barrier to guideline adherence.

Finally, we identified a number of environment-related external barriers that may compromise physicians’ confidence in their convincing power (Cabana et al., 1999). In previous studies it was reported that environment-related external barriers associated with lack of counselling materials, lack of time and lack of support staff particularly affect physician–patient communication (Cabana et al., 2000). Interventions to address these external barriers could have a beneficial effect on physicians’ self-efficacy.

Lack of outcome expectancy was also an important barrier to adherence to subfertility guidelines according to 29% of the physicians in our survey. Others reported similar rates of physicians’ lack of outcome expectancy regarding guideline adherence (Cabana et al., 1999). Feedback and audit to demonstrate the positive outcomes of guideline adherence have previously been described as useful interventions to deal with low outcome expectancy of physicians (Cabana and Kim, 2003).

Besides barriers to physician adherence to subfertility guidelines, we documented multiple factors that facilitate the ability of physicians to perform according to subfertility guidelines. For example, lack of agreement among physicians regarding the developmental process and quality of the IUI guideline was low. A plausible explanation for these findings is that physicians are more likely to accept guidelines developed by their own speciality organization than by the government or insurance companies (Cabana et al., 2001; Kasje et al., 2002). Clearly, emphasizing physician participation in guideline development and increasing awareness of guideline approval by speciality societies remains important for the improvement and sustenance of physicians’ confidence in subfertility guidelines.

The strong point of our study is its systematic approach to obtaining information on barriers to physician adherence to a subfertility guideline, using both qualitative and quantitative research methods (Grol and Wensing, 2004). We organized focus group discussions to identify potential obstacles to subfertility guideline adherence and performed an extensive questionnaire study among all Dutch physicians involved in subfertility care to quantify the prevalence and intensity of the different barriers (Grol and Wensing, 2001).

The results of our survey must be interpreted in the light of limitations regarding data collection from a self-selected sample. Physicians who lacked familiarity with the IUI guideline or who disagreed with the concept of guidelines in general or with this specific subfertility guideline may have been less likely to participate in our study. We investigated reasons for refusal to cooperate among 25% of non-responders. Since none of these physicians indicated unfamiliarity or lack of agreement as a reason for non-response and because the overall response rate of 65% was adequate, we suspect that our results have considerable validity and therefore the issue of non-response was not further addressed.

The generalizability of our findings may seem limited, because we focused on a specific subfertility guideline regarding IUI. Nevertheless, our results apply to other guidelines developed for comparable subfertility care settings, because these guidelines share rather similar developmental procedures, goals and content, target populations and target users.

Overall, our results have implications for initiatives to improve physician adherence to subfertility guidelines in order to optimize subfertility care. Whereas time and resources for quality improvement activities are generally limited, tailored interventions to address specific barriers to physician guideline adherence may be most effective and efficient (Grol, 1997; Grol and Wensing, 2001; Grol and Grimshaw, 2003). This study provides an overview of the range of barriers that prevent physician adherence to subfertility guidelines. Important barriers include physicians’ lack of self-efficacy regarding physician–patient communication and low outcome expectancy. Communication skills training for physicians and targeted education programmes for patients and physicians to facilitate shared decision-making are suggested in order to overcome barriers related to poor self-efficacy. Feedback and audit about positive outcomes of physician guideline adherence could address the lack of outcome expectancy. Further research into patients’ preferences about subfertility care is needed. Clearly, both physicians and patients play an important role in future implementation interventions to optimize subfertility care.

The authors would like to thank Dr Wim Willemsen for moderating the two focus group discussions and Annelies Pellegrino for assisting with data management. We also wish to thank the physicians who participated in the focus group interviews. The Netherlands Organisation for Health Research and Development (ZonMw) funded this project (Grant No. 945-12-012).

References

Author notes

¹Department of Obstetrics and Gynaecology and ²Centre for Quality of Care Research (WOK), Radboud University, Nijmegen Medical Centre, P.O. Box 9101, 6500 HB Nijmegen, The Netherlands