Improving prevention in primary care: evaluating the effectiveness of outreach facilitation.

BACKGROUND: Out reach facilitation is designed to promote uptake of evidence-based guidelines. There is evidence indicating that outreach facilitation can be effective in improving implementation of preventive care in GPs' offices. In this trial, we test a modified version of an outreach facilitation intervention. OBJECTIVE: To evaluate whether a comprehensive preventive intervention program using outreach facilitators improves preventive care delivery. DESIGN: Match-paired, cluster-randomized controlled trial. SETTING: Fee-for-service primary care practices in Eastern Ontario, Canada, at a time of physician shortage. PARTICIPANTS: Volunteer sample of 54 primary care practices. MAIN OUTCOME MEASURES: Mean difference between trial arms in practices' delivery of preventive manoeuvres, measured by preventive performance indices estimated from chart reviews and patient survey data. RESULTS: No difference was detected between the trial's arms for the primary outcome's overall prevention index [2.0%; 95% confidence interval (CI) -3.2 to 7.3; P = 0.44]. A small significant difference between the arms was detected for the secondary outcome's overall prevention index (2.8%; 95% CI 0.7-4.8; P = 0.01). CONCLUSION: In contrast to similar facilitation trials, this outreach facilitation program did not produce improvements in the delivery of preventive care. This lack of effect may be due to differences in the intervention and context, or the practice's limited capacity to change. Our intervention simultaneously facilitated a high number of manoeuvres, blinded facilitators and physicians to the targeted tests and had a relatively short intervention period and large number of practices assigned per facilitator. Changes in the primary care service model in Ontario at the time of the trial could have also washed out the intervention effect.

[Screening for violence against women. Validation and feasibility studies of a French screening tool]. / Dépistage de la violence faite aux femmes. Epreuves de validation et de fiabilité d'un instrument de mesure français.

OBJECTIVE: To replicate, in a Francophone community, our prior work determining the reliability and validity of the full Woman Abuse Screening Tool (WAST) and a two-item version (WAST-Short). DESIGN: Questionnaires completed by abused and nonabused women. SETTING: Two women's shelters in Francophone communities in Ontario and Quebec and participants' homes or workplaces. PARTICIPANTS: A convenience sample of 25 abused women currently residing in two women's shelters and a convenience sample of 21 women who reported they were not abused. MAIN OUTCOME MEASURES: Women's responses to French versions of the WAST, the Abuse Risk Inventory (ARI), and comfort in answering the questions were compared. Also, the reliability and validity of French versions of WAST and WAST-Short were assessed. RESULTS: Abused (n = 23) and not abused (n = 21) women were demographically similar. A strong single-factor structure that accounted for 81% of total variance in the French WAST items was identified. The French WAST was found to be highly reliable with a coefficient alpha of .95 and demonstrated construct and discriminant validity. The WAST-Short correctly classified all the nonabused women and 78.7% of the abused women. The abused women reported feeling less comfortable responding to the WAST questions than the nonabused women. CONCLUSION: The French version of the WAST demonstrated good reliability and validity and discriminated between known samples of abused and nonabused women. Even though the French WAST-Short did not perform as well as the English version, results of this study support further evaluation of the WAST for screening women in Francophone or bilingual family practice settings.

The effect of cluster randomization on sample size in prevention research.

BACKGROUND: This paper concerns the issue of cluster randomization in primary care practice intervention trials. We present information on the cluster effect of measuring the performance of various preventive maneuvers between groups of physicians based on a successful trial. We discuss the intracluster correlation coefficient of determining the required sample size and the implications for designing randomized controlled trials where groups of subjects (e.g., physicians in a group practice) are allocated at random. METHODS: We performed a cross-sectional study involving data from 46 participating practices with 106 physicians collected using self-administered questionnaires and a chart audit of 100 randomly selected charts per practice. The population was health service organizations (HSOs) located in Southern Ontario. We analyzed performance data for 13 preventive maneuvers determined by chart review and used analysis of variance to determine the intraclass correlation coefficient. An index of "up-to-datedness" was computed for each physician and practice as the number of a recommended preventive measure done divided by the number of eligible patients. An index called "inappropriateness" was computed in the same manner for the not-recommended measures. The intraclass correlation coefficients for 2 key study outcomes (up-to-datedness and inappropriateness) were also calculated and compared. RESULTS: The mean up-to-datedness score for the practices was 53.5% (95% confidence interval [CI], 51.0%-56.0%), and the mean inappropriateness score was 21.5% (95% CI, 18.1%-24.9%). The intraclass correlation for up-to-datedness was 0.0365 compared with inappropriateness at 0.1790. The intraclass correlation for preventive maneuvers ranged from 0.005 for blood pressure measurement to 0.66 for chest radiographs of smokers, and as a consequence required the sample size ranged from 20 to 42 physicians per group. CONCLUSIONS: Randomizing by practice clusters and analyzing at the level of the physician has important implications for sample size requirements. Larger intraclass correlations indicate interdependence among the physicians within a cluster; as a consequence, variability within clusters is reduced, and the required sample size increased. The key finding that many potential outcome measures perform differently in terms of the intracluster correlation reinforces the need for researchers to carefully consider the selection of outcome measures and adjust sample sizes accordingly when the unit of analysis and randomization are not the same.

Process evaluation of a tailored multifaceted approach to changing family physician practice patterns improving preventive care.

BACKGROUND: We conducted a process evaluation of a multifaceted outreach facilitation intervention to document the extent to which the intervention was implemented with fidelity. We also hoped to gain insight into how facilitation worked to improve preventive performance. METHODS: We used 5 data collection tools to evaluate the implementation of the intervention, and a combination of descriptive, quantitative, and qualitative analyses. Triangulation was used to attain a complete understanding of the quality of implementation. Twenty-two intervention practices with a total of 54 physicians participated in a randomized controlled trial that took place in Southwestern Ontario, Canada. The key measures of process were the frequency and time involved to deliver intervention components, the scope of the delivery and the utility of the components, and physician satisfaction with the intervention. RESULTS: Of the 7 components in the intervention model, prevention facilitators (PFs) visited the practice most often to deliver the audit and feedback, consensus building, and reminder system components. All the study practices received preventive performance audit and feedback, achieved consensus on a plan for improvement, and implemented a reminder system. Ninety percent of the practices implemented a customized flow sheet, and 10% used a computerized reminder system. Ninety-five percent of the intervention practices wanted critically appraised evidence for prevention, 82% participated in a workshop with opinion leaders in preventive care, and 100% received patient education materials in a binder. Content analysis of the physician interviews and bivariate analysis of physician self-reported changes between intervention and control group physicians revealed that the audit and feedback, consensus building, and development of reminder systems were the key intervention components. Ninety-five percent of the physicians were either satisfied or very satisfied with the intervention, and 90% would have been willing to have the PF continue working with their practice. CONCLUSIONS: Primary care practices in Ontario can implement significant changes in their practice environments that will improve preventive care activity with the assistance of a facilitator. The main components for creating change are audit and feedback of preventive performance, achieving consensus on a plan for improvement, and implementing a reminder system.

Evidence to action: a tailored multifaceted approach to changing family physician practice patterns and improving preventive care.

BACKGROUND: Although there is much room for improvement in the performance of recommended preventive manoeuvres, many inappropriate preventive interventions are being done. We evaluated a multifaceted intervention, delivered by nurses trained in prevention facilitation, to improve prevention in primary care. METHODS: Forty-six health service organizations (HSOs) were recruited from 100 sites in Ontario. After baseline data were collected, we randomly assigned the practices to either an 18-month (July 1997 to December 1998) multifaceted intervention delivered by 1 of 3 nurse facilitators (23 practices) or no intervention (23 practices). The unit of intervention and analysis was the medical practice. The outcome measure was an overall index of preventive performance, which was calculated as the proportion of eligible patients who received 8 recommended preventive manoeuvres less the proportion of eligible patients who received 5 inappropriate preventive manoeuvres. RESULTS: One HSO, in the intervention group, was lost to follow-up. Before the intervention, the index of preventive performance was similar for the intervention and control groups (31.9% [95% confidence interval (CI) 27.3%-36.5%] and 32.1% [95% CI 27.2%-37.0%] respectively). At follow-up the corresponding values were 43.2% (95% CI 38.4%-48.0%) and 31.9% (95% CI 26.8%-37.0%), for an absolute improvement in the intervention group of 11.5% (p < 0.001). The mean proportion of eligible patients who received the recommended manoeuvres was 62.3% (95% CI 58.2%-66.4%) in the intervention group, as compared with 57.4% (95% CI 54.1%-60.7%) in the control group, for an absolute improvement of 7.2% (p = 0.008). The corresponding values for the inappropriate manoeuvres were 19.1% (95% CI 15.6%-22.6%) and 25.5% (95% CI 20.0%-31.0%), for an absolute improvement of 4.4% (p = 0.019). INTERPRETATION: The tailored multifaceted intervention delivered by nurse facilitators was effective in modifying physician practice patterns and significantly improved preventive care performance.

Warfarin for atrial fibrillation. The patient's perspective.

OBJECTIVE: To determine the minimal clinically important difference (MCID) of warfarin therapy for the treatment of nonvalvular atrial fibrillation from the perspective of patients using 2 different elicitation methods. DESIGN: All patients completed 2 face-to-face interviews, which were 2 weeks apart. For each interview, they were randomized to receive 1 of 2 elicitation methods: ping-ponging or starting at the known efficacy. SETTING: The practices of 2 university-affiliated family medicine centers (8 physicians each), 14 community-based family physicians, and 2 cardiologists. PATIENTS: Sixty-four patients with nonvalvular atrial fibrillation who were initiated with warfarin therapy at least 3 months before the study. INTERVENTION: During each interview, the patients' MCIDs were determined by using (1) a pictorial flip chart to describe atrial fibrillation; the consequences of a minor stroke, a major stroke, and a major bleeding episode; the chance of stroke if not taking warfarin; the chance of a major bleeding episode if taking warfarin; examples of the inconvenience, minor side effects, and costs of warfarin therapy; and then (2) 1 of the 2 elicitation methods to determine their MCIDs (the smallest reduction in stroke risk at which the patients were willing to take warfarin). Patients' knowledge of their stroke risk, acceptability of the interview process, and factors determining their preferences were also assessed. MAIN RESULTS: Given a baseline risk of having a stroke in the next 2 years, if not taking warfarin, of 10 of 100, the mean MCID was 2.01 of 100 (95% confidence interval, 1.60-2.42). Fifty-two percent of the patients would take warfarin for an absolute decrease in stroke risk of 1% over 2 years. Before eliciting their MCIDs, patients showed poor knowledge of their stroke risk, which improved afterward. The interview process was well accepted by the patients. The MCID using the ping-ponging elicitation method was 1.015 of 100 smaller compared with use of the starting at the known efficacy method (P = .01). CONCLUSIONS: We were able to determine the MCID of warfarin therapy for the prevention of stroke from the perspective of patients with nonvalvular atrial fibrillation. Their MCIDs were much smaller than those that have been implied by some experts and clinicians. The interview process, using the flip chart approach, appeared to improve the patients' knowledge of their disease and its consequences and treatment. The method used to elicit the patients' MCIDs can have a clinically important effect on patient responses. The method used in our study can be generalized to other conditions and, thus, could be helpful in 3 ways: (1) from a clinical decision-making perspective, it could facilitate patient-physician communication; (2) it could clarify the patient perspective when interpreting the results of previously completed trials; and (3) it could be used to derive more clinically relevant sample sizes for randomized treatment trials.

Depression in primary care. Why do we miss the diagnosis?

Depressed patients are often undiagnosed, misdiagnosed, or underdiagnosed. Is this because family physicians are trained mainly to treat somatic complaints? Are patients reluctant to accept psychological causes for their physical symptoms? High volume of patients and short visits make it difficult for doctors to recognize depression. We propose strategies for identifying depressed patients in primary care.

Family practice informatics: research issues in computerized medical records.

There are unique features of family and general practice which lead to unique issues in medical informatics for family physicians. The nature of practice in office based community settings and the discipline of dealing with all ages, sexes, and health conditions over the lifetime of a patient and his/her family lead to models of the thinking that are different from those used in most other specialties. Research is urgently needed to verify the models of thinking that physicians use during patient care encounters and the associated nomenclatures and classifications which support them. User interfaces need to be optimized for accuracy and speed. Standards for medical records computing in family practice need testing and validation.

Does Antihypertensive Therapy Need to be Life-Long?

The author reviews the evidence for and against decreasing or discontinuing antihypertensive therapy on known hypertensive patients once their blood pressure has been brought under control. The evidence supports a trial of decreasing and, in many cases, discontinuing antihypertensive medication. Although there are no established protocols for cessation of therapy, the author discusses suggestions in the literature.

Enuresis: are we using the optimal treatment?

A multitude of treatments are currently being used for enuresis. This article reviews the scientific evidence supporting various treatments and indicating their effectiveness. Recent evidence demonstrating the significant improvement of psychological well-being as a result of treatment indicates that treatment should be started early. The literature shows that 48% of parents do not seek medical help for children with enuresis. They tend to use ineffective and sometimes damaging treatment, such as punishment. The authors conclude that the modern enuretic alarm is relatively safe, painless, easy to use, and is the most effective treatment available.

The case for small rural hospital obstetrics.

Small obstetrical units can offer a very competitive quality of service as compared to large centres. They achieved this quality of service by transferring all high-risk deliveries to specialized regional centres with neonatal intensive care facilities (perinatal regionalization). The significant decrease in perinatal mortality over the past 20 years is a result of the superior care of low birth-weight infants, available in these large centres. It is not surprising that some people might project this superior performance to all weight categories. A review of the literature, however, shows that overall, small (level I) hospitals are as safe as larger centres. The authors discuss the means by which small, and often isolated, centres do as well as better equipped and more extensive centres with larger staffs.

Referral pattern and rate of intervention in a small rural obstetrical practice.

This article describes an obstetrical chart review that covered a seven-year period from April 30, 1978 to March 31, 1984. This review includes both pre- and intra partum transfers and involves two main topics: referral pattern and rate of intervention during labour. The practice was involved in 60% of all pregnancies (344) of the catchment area. Seventeen per cent were transferred in their pre-partum period, and 11% were transferred intra partum. The rates of intervention for forceps, induction, episiotomy and analgesia are presented, and they confirm the non-interventionist attitude of general practice obstetricians.