[Meaningful interventions for peripheral arterial disease: Stopping primary referral to the vascular surgeon and angioplasty]. / Zinvolle interventies bij perifeer vaatlijden.

Supervised exercise therapy with lifestyle counseling by a specialized physical therapist has replaced endovascular or open revascularization as primary treatment for intermittent claudication for most patients in the Netherlands. Succesful implementation of this guideline-recommended treatment strategy was achieved by warranting availability and quality of care by organizing care around a national network. Further reductions in unnecessary interventions in this population may be achieved when general practitioners and medical specialists collaborate, for instance by making in-hospital vascular diagnostic laboratories accessible for primary care. The implementation of similar interventions for other non-communicable chronic diseases such as chronic obstructive pulmonary disease or coronary heart disease in standard care is lacking, despite a solid evidence base and guideline recommendations.

Learning Conversations with Trainees: An Undervalued but Useful EBM Learning Opportunity for Clinical Supervisors.

Phenomenon: Supervisors and trainees can learn skills related to evidence-based medicine from each other in the workplace by collaborating and interacting, in this way benefiting from each other's strengths. This study explores supervisors' perceptions of how they currently learn evidence-based medicine by engaging in learning conversations with their trainee. Approach: Semi-structured, video-stimulated elicitation interviews were held with twenty-two Dutch and Belgian supervisors in general practice. Supervisors were shown fragments of their video-recorded learning conversations, allowing them to reflect. Recorded interviews were analyzed using a grounded theory-based approach.Findings: Supervisors did not immediately perceive workplace learning conversations as an opportunity to learn evidence-based medicine from their trainee. They mostly saw these conversations as a learning opportunity for trainees and a chance to maintain the quality of care within their practice. Nevertheless, during the interviews, supervisors did acknowledge that learning conversations help them to gain up-to-date knowledge and search skills or more awareness of their own knowledge or gaps in their knowledge. Not identified as a learning outcome was how to apply evidence-based medicine within a clinical practice by combining evidence with clinical expertise and the patient's preferences. Insights: Supervisors acknowledge that they learn elements of the three aspects of evidence-based medicine by having learning conversations with their trainee, but they currently see this as secondary to the trainee's learning process. Emphasizing opportunities for bidirectional learning could improve learning of evidence-based medicine during workplace learning conversations.

GP trainees' perceptions on learning EBM using conversations in the workplace: a video-stimulated interview study.

BACKGROUND: To be able to practice evidence-based medicine (EBM) when making decisions for individual patients, it is important to learn how to combine the best available evidence with the patient's preferences and the physician's clinical expertise. In general practice training, these skills can be learned at the workplace using learning conversations: meetings between the supervising general practitioner (GP) and GP trainee to discuss medical practice, selected topics or professional performance. This study aimed to give insight into the perceptions of GP trainees on their EBM learning processes during learning conversations. METHODS: We held semi-structured video-stimulated elicitation interviews (n = 22) with GP trainees affiliated to GP training institutes in the Netherlands and Belgium. GP trainees were shown fragments of their learning conversations, enabling reflection during the interview. Taking an inductive approach, interview recordings were transcribed verbatim and analysed with NVivo software. RESULTS: GP trainees perceived learning conversations as useful for learning and discussing EBM. Multiple EBM learning activities were identified, such as discussing evidence together, relating evidence to cases in daily practice and discussing the supervisor's experience and the specific local context in the light of what the evidence recommends. However, for learning to occur, trainees need and expect specific behaviour, both from their supervisors and themselves. Supervisors should supply well-substantiated answers that are applicable in practice and give the trainee confirmation. In turn, the trainee needs to prepare well in order to ask focused, in-depth questions. A safe space allowing equal and open discussion between trainee and supervisor is perceived as an essential context for optimal EBM learning. CONCLUSIONS: Our findings show that trainees find learning conversations useful for EBM learning in general practice. To bring EBM learning to its full potential, attention should be paid to optimising the behavioural and contextual factors found relevant to enhancing EBM learning.

Applying evidence-based medicine in general practice: a video-stimulated interview study on workplace-based observation.

BACKGROUND: Evidence-based medicine (EBM) in general practice involves applying a complex combination of best-available evidence, the patient's preferences and the general practitioner's (GP) clinical expertise in decision-making. GPs and GP trainees learn how to apply EBM informally by observing each other's consultations, as well as through more deliberative forms of workplace-based learning. This study aims to gain insight into workplace-based EBM learning by investigating the extent to which GP supervisors and trainees recognise each other's EBM behaviour through observation, and by identifying aspects that influence their recognition. METHODS: We conducted a qualitative multicentre study based on video-stimulated recall interviews (VSI) of paired GP supervisors and GP trainees affiliated with GP training institutes in Belgium and the Netherlands. The GP pairs (n = 22) were shown fragments of their own and their partner's consultations and were asked to elucidate their own EBM considerations and the ones they recognised in their partner's actions. The interview recordings were transcribed verbatim and analysed with NVivo. By comparing pairs who recognised each other's considerations well with those who did not, we developed a model describing the aspects that influence the observer's recognition of an actor's EBM behaviour. RESULTS: Overall, there was moderate similarity between an actor's EBM behaviour and the observer's recognition of it. Aspects that negatively influence recognition are often observer-related. Observers tend to be judgemental, give unsolicited comments on how they would act themselves and are more concerned with the trainee-supervisor relationship than objective observation. There was less recognition when actors used implicit reasoning, such as mindlines (internalised, collectively reinforced tacit guidelines). Pair-related aspects also played a role: previous discussion of a specific topic or EBM decision-making generally enhanced recognition. Consultation-specific aspects played only a marginal role. CONCLUSIONS: GP trainees and supervisors do not fully recognise EBM behaviour through observing each other's consultations. To improve recognition of EBM behaviour and thus benefit from informal observational learning, observers need to be aware of automatic judgements that they make. Creating explicit learning moments in which EBM decision-making is discussed, can improve shared knowledge and can also be useful to unveil tacit knowledge derived from mindlines.

Serum biomarkers in patients suspected of transient ischaemic attack in primary care: a diagnostic accuracy study.

OBJECTIVE: The diagnosis of transient ischaemic attack (TIA) based on symptoms and signs can be challenging and would greatly benefit from a rapid serum biomarker of brain ischaemia. We aimed to quantify the added diagnostic value of serum biomarkers in patients suspected of TIA beyond symptoms and signs. METHODS: This is a cross-sectional diagnostic accuracy study with a 6-month follow-up period. Participants were patients suspected of TIA by the general practitioner (GP) in whom a blood sample could be collected within 72 hours from symptom onset. A research nurse visited the participant for the blood sample and a standardised interview. The GP referred participants to the regional TIA service. An expert panel of three neurologists classified cases as TIA, minor stroke or any other diagnosis, based on all available diagnostic information including the GP's and neurologist's correspondence and the follow-up period. We used multivariable logistic regression analyses to quantify the diagnostic accuracy of clinical predictors and the improvement of accuracy by seven biomarkers (NR2, NR2 antibodies, PARK7, NDKA, UFD1, B-FABP and H-FABP). RESULTS: 206 patients suspected of TIA participated, of whom 126 (61.2%) were diagnosed with TIA (n=104) or minor stroke (n=22) by the expert panel. The median time from symptom onset to the blood sample collection was 48.0 (IQR 28.3-56.8) hours. None of the seven biomarkers had discriminative value in the diagnosis of TIA, with C-statistics ranging from 0.45 to 0.58. The final multivariable model (C-statistic 0.83 (0.78-0.89)) consisted of eight clinical predictors of TIA/minor stroke: increasing age, a history of coronary artery disease, sudden onset of symptoms, occurrence of symptoms in full intensity, dysarthria, no history of migraine, absence of loss of consciousness and absence of headache. Addition of the individual biomarkers did not further increase the C-statistics. CONCLUSIONS: Currently available blood biomarkers have no added diagnostic value in suspected TIA. TRIAL REGISTRATION NUMBER: NCT01954329.

Candidate Biomarkers for the Diagnosis of Transient Ischemic Attack: A Systematic Review.

BACKGROUND AND PURPOSE: A rapid serum biomarker that confirms or rules out a transient ischemic attack (TIA) would be of great value in clinical practice. We aimed to systematically review current evidence for the diagnostic accuracy of blood biomarkers in the early diagnosis of TIA. METHODS: This is a systematic review with quality appraisal of individual studies using the QUADAS-2 tool. MEDLINE and EMBASE databases were searched up to May 1, 2017, to select primary diagnostic accuracy studies evaluating potential biomarkers in blood for the diagnosis of TIA or ischemic stroke. RESULTS: Of 4,215 studies retrieved, 78 met our eligibility criteria. Forty-five studies restricted their population to ischemic stroke patients, 32 included both TIA and ischemic stroke patients, and only one study was restricted to TIA patients. In total 62/78 (79.5%) studies had a case-control design comparing TIA or stroke patients with healthy subjects. Overall, 125 single biomarkers and 5 biomarker panels were studied, with a median number of participants per study of 92.0 (interquartile range 44.8-144.5), varying from 8 to 915. Sufficient information to extract 2 × 2 tables was available for 35 (44.9%) articles, and for 60 (48.0 %) biomarkers. Several markers, such as NR2A/B (antibodies), Parkinson 7, nucleoside diphosphate kinase A, ubiquitin fusion degradation protein-1, and heart-type fatty acid binding protein, have shown moderate to high diagnostic accuracy in multiple studies. CONCLUSIONS: Although the methodological quality of studies evaluating biomarkers of brain ischemia was poor, several biomarkers have shown the potential to detect transient brain ischemia in an early phase. Diagnostic accuracy studies in suspected cases of TIA are needed to determine their true clinical value.

Determinants of Patient Delay in Transient Ischemic Attack.

INTRODUCTION: Early diagnosis and stroke preventive treatment in patients with transient ischemic attack (TIA) are crucial, but hampered by delayed reporting of symptoms. Previous studies on causes of patient delay provided inconsistent results. We aimed to assess determinants of patient delay among patients with symptoms suggestive of TIA. METHODS: We interviewed participants referred by their general practitioner to an outpatient TIA clinic within 72 h from symptom onset. We determined (i) the exact time from symptom onset to the first contact with a medical service (patient delay); (ii) demographic and clinical characteristics; (iii) patient's initial perception, and reaction to symptoms; and (iv) patient's knowledge about TIA. We used multivariable linear regression to identify determinants of patient delay. RESULTS: We interviewed 202 suspected TIA patients (mean age 67.7 (SD 13.7) years, 111 (55.0%) male), of whom 123 (60.9%) received a definite diagnosis of TIA or minor stroke. Median patient delay was 1.5 (interquartile range 0.4-14.6) hours. Of all patients, 119 (58.9%) considered a TIA (or stroke) as the cause of their symptoms. Among them, 30 (25.2%) thought it was a medical emergency, while of the 83 not considering TIA as the cause of symptoms 38 (45.8%) thought of a medical emergency. Independently related to increased delay were (i) symptom onset out of hours, (ii) absence of dysarthria, (iii) being unaware that TIA requires urgent treatment, (iv) not considering the event an emergency, and (v) knowledge of TIA symptoms. Results for patients with a definite diagnosis of TIA/minor stroke were similar to those with alternative diagnoses. CONCLUSION: Patients still tend to wait till office hours to report TIA symptoms. Speech difficulties, and specifically dysarthria, are related to shorter delay. To reduce patient delay, awareness of TIA symptoms should increase and more importantly lay people should be educated to consider a TIA an emergency.

Understanding the Broker Role of Clinician-Scientists: A Realist Review on How They Link Research and Practice.

PURPOSE: Clinician-scientists are said to be well placed to connect research and practice, but their broker role has been underexplored. This review sought to gain an understanding of the broker role of clinician-scientists. METHOD: The authors conducted a realist review to describe context-mechanism-outcome (CMO) configurations associated with the broker role of clinician-scientists. CINAHL, PubMed, PsycINFO, Web of Science, and Embase were searched between May and August 2017. Data were analyzed qualitatively; data synthesis focused on assembling CMO configurations. RESULTS: Of an initial 2,241 articles, 9 were included in the final review. Included papers show that clinician-scientists, in their broker role, achieve 2 organizational-level outcomes: an increased volume of clinically relevant, research, and increased evidence application to improve care. They also achieve the individual-level outcome of professional development as a researcher, clinician, and broker. Multidimensional skills and management support are necessary context factors. Mechanisms that contribute to outcomes include balancing economic and scientific interests and performing boundary-crossing activities. Four CMO configurations by which clinician-scientists achieve outcomes in brokering a connection between research and practice were identified. Useful program theories for explaining these are boundary crossing, social network, communities of practice, and diffusion of innovation theory. CONCLUSIONS: The mechanisms found may provide insight for interventions aiming to support clinician-scientists in their broker role. The authors expect that if more attention is paid to learning multidimensional skills and management support for the broker role is strengthened, stronger links between research and practice could be forged.

Patient delay in TIA: a systematic review.

BACKGROUND: Patients who suffer a transient ischemic attack (TIA) have a high short-term risk of developing ischemic stroke, notably within the first 48 h. Timely diagnosis and urgent preventive treatment substantially reduce this risk. We conducted a systemic review to quantify patient delay in patients with (suspected) TIA, and assess determinants related to such delay. METHODS: A systematic review using MEDLINE and EMBASE databases up to March 2017 to identify studies reporting the time from onset of TIA symptoms to seeking medical help. RESULTS: We identified nine studies providing data on patient delay, published between 2006 and 2016, with 7/9 studies originating from the United Kingdom (UK). In total 1103 time-defined TIA patients (no remaining symptoms > 24 h), and 896 patients with a minor stroke (i.e., mild remaining symptoms > 24 h) were included (49.1% men, mean age 72.2 years). Patient's delay of more than 24 h was reported in 33.1-44.4% of TIA patients, with comparable proportions for minor stroke patients. Delays were on average shorter in patients interviewed at the emergency department than among patients seen at TIA outpatient clinics. Univariably associated with a shorter delay were (1) a longer duration of symptoms, (2) motor symptoms, (3) a higher ABCD2 score, and (4) correct patient's recognition as possible ischemic cerebrovascular event. CONCLUSIONS: More than a third of patients experiencing a TIA delays medical attention for more than a day, thus critically extending the initiation of stroke preventive treatment. There still seems to be insufficient awareness among lay people that symptoms suggestive of TIA should be considered as an emergency. Additional data and multivariable analyses are needed to define main determinants of patient delay.

The use of evidence during group meetings of Dutch general practitioners.

BACKGROUND AND OBJECTIVES: For Evidence Based Medicine (EBM) learning journal clubs are recommended, but these are not common practice. How participants discuss and learn about applying evidence in other group meetings is unknown. We examined different types of group meetings and explored the use of, and discussions about, clinical research evidence. METHODS: A mixed-methods study design was adopted. After distribution and analysis of a questionnaire about types of group meetings, interviews were conducted to better understand the most frequently occurring type. RESULTS: GPs have different types of meetings, but the most common group meetings where evidence wass discussed were so called quality circles, i.e. pharmacotherapy audit meetings in which GPs discuss drug prescription figures or preferred treatment together with pharmacists. Interviews showed that the source of evidence used mostly are the recommendations in the national GP guidelines. The underlying evidence or new research did not play an important role in the discussions. CONCLUSIONS: Quality circles seem to be more goal-oriented than learning oriented. Learning discussions about controversies in clinical research or about the integration of evidence, patient values and clinical expertise occurred infrequently. To harvest the potential value of group meetings for EBM learning, quality circles in their present design are not optimal.

The Utrecht questionnaire (U-CEP) measuring knowledge on clinical epidemiology proved to be valid.

OBJECTIVES: Knowledge on clinical epidemiology is crucial to practice evidence-based medicine. We describe the development and validation of the Utrecht questionnaire on knowledge on Clinical epidemiology for Evidence-based Practice (U-CEP); an assessment tool to be used in the training of clinicians. STUDY DESIGN AND SETTING: The U-CEP was developed in two formats: two sets of 25 questions and a combined set of 50. The validation was performed among postgraduate general practice (GP) trainees, hospital trainees, GP supervisors, and experts. Internal consistency, internal reliability (item-total correlation), item discrimination index, item difficulty, content validity, construct validity, responsiveness, test-retest reliability, and feasibility were assessed. The questionnaire was externally validated. RESULTS: Internal consistency was good with a Cronbach alpha of 0.8. The median item-total correlation and mean item discrimination index were satisfactory. Both sets were perceived as relevant to clinical practice. Construct validity was good. Both sets were responsive but failed on test-retest reliability. One set took 24 minutes and the other 33 minutes to complete, on average. External GP trainees had comparable results. CONCLUSION: The U-CEP is a valid questionnaire to assess knowledge on clinical epidemiology, which is a prerequisite for practicing evidence-based medicine in daily clinical practice.

Development and validation of a new instrument measuring guideline adherence in clinical practice.

BACKGROUND: Education in evidence-based medicine (EBM) is an important part of the postgraduate training of GPs. Evaluation of its effect on EBM behaviour in daily clinical practice is difficult and instruments are scarce. Working in accordance with guidelines is considered as one of the key indicators of EBM behaviour. OBJECTIVE: To develop and validate an instrument assessing guideline adherence of GP trainees in clinical practice. METHODS: We developed an instrument that assesses guideline adherence, taking conscious deviation into account. The instrument assesses guideline adherence on 59 different management decisions (diagnosis N = 17, therapy N = 20, referral N = 22) for 23 conditions as described in 27 different clinical practice guidelines. We validated this instrument using performance data as collected by third-year GP trainees on three important properties: validity, reliability and feasibility. RESULTS: Performance data were collected by 76 GP trainees on 12106 patient consultations with 12587 different reasons for encounter. Overall, guideline adherence was 82% (95% confidence interval 77-88%). The significant correlation with the national GP knowledge test (r 0.33, P 0.004) showed the instrument to be a valid instrument. Interrater reliabilities (intraclass correlation coefficient) varied between moderate and excellent (0.64-1.00, P < 0.001). The instrument proved feasible with coverage of 24% (N = 3082) of reasons for encounter presented to GP trainees and a mean and median time of 1 minute to score a patient consultation. CONCLUSION: This instrument proved valid, reliable and feasible to assess guideline adherence among trainees in the clinical primary care setting.

General practice trainees' information searching strategies for clinical queries encountered in daily practice.

BACKGROUND: Earlier studies have shown that clinical queries are common among doctors. Data on the information-seeking behaviour of general practice (GP) trainees are scarce though, and numbers studied are small. OBJECTIVE: The objective of this study was to determine how often and how GP trainees search for answers to clinical queries encountered in daily clinical practice. METHODS: Third-year GP trainees kept logs on all patient contacts for eight consecutive practice days. Information was obtained on patient contacts (description), clinical queries (frequency, type), seeking behaviour (frequency, moment, reason not to search, resources used, duration of search) and answers (frequency, impact). Descriptive analyses were performed; frequencies and percentages were computed. We calculated the number of clinical queries per patient, the number of searches per query and the number of answers per search. RESULTS: Seventy-six trainees reported 1533 clinical queries about 7300 patients presenting 7619 complaints [mean of 0.2 queries per patient, standard deviation (SD) 0.1]. For most of the queries trainees pursued an answer (mean of 0.8 per query, SD 0.2), mostly during consultation (61% of searches), and frequently retrieved answers (mean of 0.8 per search, SD 0.17) they reported to improve clinical decision making in 26%. Most common resources were colleagues or supervisors (28%), and national GP guidelines (26%). The median duration of a search was 4 minutes (interquartile range 3). CONCLUSION: GP trainees have one clinical query per five patients. They often attempted to find answers and reported to succeed in most of the searches, primarily by consulting supervisors or colleagues and national GP guidelines.

Lifetime endogenous estrogen exposure and electrocardiographic frontal T axis changes in postmenopausal women.

OBJECTIVE: The protective effect of endogenous estrogens in cardiovascular disease may in part originate from effects of circulating estrogens on the electrophysiological properties of the myocardium. The aim of this study was to investigate the relation between reproductive factors and the electrocardiographic frontal T axis in postmenopausal women. DESIGN: Cohort study. SETTING: The study was conducted at the University Medical Center Utrecht. PATIENTS: In total, 998 postmenopausal women were included. MAIN OUTCOMES: Information of women's reproductive life was obtained by a questionnaire. Electrocardiographic frontal T axes were categorized as normal (25-65 degrees) or abnormal (-180 degrees to 24 degrees and 66-180 degrees). Logistic regression analysis was used to assess the relationship between reproductive factors and the frontal T axis. Moreover, the effect of the lifetime cumulative number of menstrual cycles, a composite measure of all reproductive factors, on the frontal T axis was investigated. RESULTS: The mean age was 66.0 (+/-5.6) years and 15.3% had T-axes abnormalities. Later menopausal age decreased the risk on frontal T-axis abnormalities; the multivariable adjusted odds ratio was 0.97 (95% CI: 0.94-0.99) per year increasing menopause. For the lifetime cumulative number of menstrual cycles the age-adjusted odds ratio was 0.84 (95% CI: 0.75-0.99) per 100 menstrual cycles increase. CONCLUSIONS: Later age at menopause and increasing lifetime cumulative number of menstrual cycles decreased the risk on frontal T-axis changes. This supports the view that estrogens may protect against ventricular repolarization disturbances.

Ankle brachial index measurement in primary care: are we doing it right?

BACKGROUND: The reference standard for diagnosing peripheral arterial disease in primary care is the ankle brachial index (ABI). Various methods to measure ankle and brachial blood pressures and to calculate the index are described. AIM: To compare the ABI measurements performed in primary care with those performed in the vascular laboratory. Furthermore, an inventory was made of methods used to determine the ABI in primary care. DESIGN OF STUDY: Cross-sectional study. SETTING: Primary care practice and outpatient clinic. METHOD: Consecutive patients suspected of peripheral arterial disease based on ABI assessment in primary care practices were included. The ABI measurements were repeated in the vascular laboratory. Referring GPs were interviewed about method of measurement and calculation of the index. From each patient the leg with the lower ABI was used for analysis. RESULTS: Ninety-nine patients of 45 primary care practices with a mean ABI of 0.80 (standard deviation [SD] = 0.27) were included. The mean ABI as measured in the vascular laboratory was 0.82 (SD = 0.26). A Bland-Altman plot demonstrated great variability between ABI measurements in primary care practice and the vascular laboratory. Both method of blood pressure measurements and method of calculating the ABI differed greatly between primary care practices. CONCLUSION: This study demonstrates that the ABI is often not correctly determined in primary care practice. This phenomenon seems to be due to inaccurate methods for both blood pressure measurements and calculation of the index. A guideline for determining the ABI with a hand-held Doppler, and a training programme seem necessary.

High dietary menaquinone intake is associated with reduced coronary calcification.

BACKGROUND: Dietary vitamin K is thought to decrease risk of cardiovascular disease by reducing coronary calcification, but inconsistent results are reported. This may be due to different effects of vitamin K(1) (phylloquinone) and vitamin K(2) (menaquinone, MK), but few studies included both. METHODS: We investigated the association of intake of phylloquinone and menaquinone, including its subtypes (MK4-MK10), with coronary calcification in a cross-sectional study among 564 post-menopausal women. Phylloquinone and menaquinone intake was estimated using a food-frequency questionnaire. RESULTS: Sixty-two percent (n=360) of the women had coronary calcification based on 1.5-mm thick slices. Phylloquinone intake was not associated with coronary calcification with a relative risk (RR) of 1.17 (95%-confidence interval: 0.96-1.42; p(trend)=0.11) of the highest versus lowest quartile. Menaquinone intake was associated with decreased coronary calcification with an RR of 0.80 (95%-CI: 0.65-0.98; p(trend)=0.03). CONCLUSION: This study shows that high dietary menaquinone intake, but probably not phylloquinone, is associated with reduced coronary calcification. Adequate menaquinone intakes could therefore be important to prevent cardiovascular disease.

Reproductive factors, metabolic factors, and coronary artery calcification in older women.

OBJECTIVE: Metabolic disturbances may explain the increased cardiovascular risk associated with reproductive factors. This cohort study investigated the relationship between reproductive factors and coronary artery calcification in elderly women and whether this relationship could be explained by metabolic disturbances. DESIGN: In total, 568 postmenopausal women were included in this cross-sectional study. Information about the women's reproductive life was obtained by a questionnaire. Metabolic factors were measured during a single visit. Coronary artery calcification was assessed with a multislice computed tomography scanner and dichotomized as absent or present. Logistic regression analysis was used to assess the relationship between reproductive factors and coronary artery calcification. Crude and multivariate adjusted odds ratios (ORs) with corresponding 95% confidence intervals (CIs) were calculated. In addition, ORs were adjusted for several metabolic and cardiovascular risk factors. RESULTS: The mean age was 66.9 (+/- 5.5) years. Women with a history of irregular menstrual cycle lengths, as opposed to women with a history of regular menstrual cycles (26-30 d), had an increased risk of coronary artery calcification; multivariate-adjusted OR = 2.73 (95%CI: 1.24-5.98). Four or more pregnancies, compared with never pregnant, yielded an multivariate-adjusted OR of 1.89 (95% CI: 1.00-3.58). Having four children or more, compared with having no children, yielded a multivariate-adjusted OR of 1.97 (95% CI: 1.00-3.89). Adjustment for metabolic factors and other cardiovascular risk factors did not fully explain theses relationships. CONCLUSION: Multigravidity (more than four pregnancies), multiparity (more than four births), and irregular menstrual cycle lengths were related to an increased risk of coronary artery disease. These associations could not be explained by metabolic abnormalities.