The role of 'confounding by indication' in assessing the effect of quality of care on disease outcomes in general practice: results of a case-control study.

BACKGROUND: In quality of care research, limited information is found on the relationship between quality of care and disease outcomes. This case-control study was conducted with the aim to assess the effect of guideline adherence for stroke prevention on the occurrence of stroke in general practice. We report on the problems related to a variant of confounding by indication, that may be common in quality of care studies. METHODS: Stroke patients (cases) and controls were recruited from the general practitioner's (GP) patient register, and an expert panel assessed the quality of care of cases and controls using guideline-based review criteria. RESULTS: A total of 86 patients was assessed. Compared to patients without shortcomings in preventive care, patients who received sub-optimal care appeared to have a lower risk of experiencing a stroke (OR 0.60; 95% CI 0.24 to 1.53). This result was partly explained by the presence of risk factors (6.1 per cases, 4.4 per control), as reflected by the finding that the OR came much closer to 1.00 after adjustment for the number of risk factors (OR 0.82; 95% CI 0.29 to 2.30). Patients with more risk factors for stroke had a lower risk of sub-optimal care (OR for the number of risk factors present 0.76; 95% CI 0.61 to 0.94). This finding represents a variant of 'confounding by indication', which could not be fully adjusted for due to incomplete information on risk factors for stroke. CONCLUSIONS: At present, inaccurate recording of patient and risk factor information by GPs seriously limits the potential use of a case-control method to assess the effect of guideline adherence on disease outcome in general practice. We conclude that studies on the effect of quality of care on disease outcomes, like other observational studies of intended treatment effect, should be designed and performed such that confounding by indication is minimized.

Quality of care in stroke prevention: results of an audit study among general practitioners.

BACKGROUND: In identifying opportunities to improve the quality of stroke prevention in general practice, insight in areas of suboptimal care is essential. This study investigated the quality of care in stroke prevention in general practice and its relation to the occurrence of stroke. METHODS: Retrospective case-based audit with guideline-based review criteria and final judgment of suboptimal care by an expert panel. RESULTS: A total of 292 stroke patients were identified through stroke registers of two main referral hospitals for stroke in Rotterdam. The general practitioners (GPs) (n = 95) of these patients were approached. The overall response rate from GPs was 81%, and a total of 193 patients from 77 GPs were included in the study. Data on the process of care at patient level were collected by chart review and by structured interviews with GPs during site visits. All cases were presented to a six-member panel of GPs and neurologists. In 44% of the cases, suboptimal care was identified (31% judged as possibly or likely failing to prevent stroke). Of the total number of identified shortcomings, 52% was related to inadequate hypertension control, particularly lack of follow-up after established hypertension. Another 17% of identified shortcomings concerned inadequate cardiovascular risk assessment. CONCLUSIONS: A substantial number of shortcomings in care, particularly in the domain of hypertension control and the assessment of patient's risk profiles for cardiovascular disease (CVD), were identified. This study suggests that improving preventive care delivery in general practice could reduce the occurrence of stroke.

Perceptions of cardiovascular risk among patients with hypertension or diabetes.

We aimed to examine risk perceptions among patients at moderate to high cardiovascular risk. A questionnaire about perceived absolute risk of myocardial infarction and stroke was sent to 2424 patients with hypertension or diabetes. Response rate was 86.3% and 1557 patients without atherosclerotic disease were included. Actual cardiovascular risk was calculated by using Framingham risk functions. A total of 363 (23.3%) of the 1557 patients did not provide any risk estimates and these were particularly older patients, patients with a lower educational level, and patients reporting no alcohol consumption. The remaining 1194 patients tended to overestimate their risk. In 42.3% (497/1174) and 46.8% (541/1155) of the cases, patients overestimated their actual 10-year risk for myocardial infarction and stroke, respectively, by more than 20%. Older age, smoking, familial history of cardiovascular disease (CVD), and actual absolute risk predicted higher levels of perceived absolute risk. Male sex, higher scores for an internal health locus of control, lower scores for a physician locus of control, and self-rated excellent or (very) good health were positively related to higher accuracy. In conclusion, patients showed inadequate perceptions of their absolute risk of cardiovascular events and physicians should thus provide greater information about absolute risk when offering preventive therapy.

Organisational determinants of cardiovascular prevention in general practice.

OBJECTIVE: To assess organisational determinants in the prevention of cardiovascular disease. DESIGN: A cross-sectional study. SETTING AND SUBJECTS: 130 general practices in The Netherlands. Data were collected using questionnaires. A causal model was designed and analysed by path analysis. MAIN OUTCOME MEASURES: Relationships between preventive activities, practice management and practice characteristics. RESULTS: Important differences between adequacy of equipment and practice organisation were found. Record-keeping was significantly better when working experience of the GPs was less than 15 years, when the practice consisted only of female GPs, and when the practice had written protocols for cardiovascular disease management and the staff held regular scheduled meetings (teamwork). Teamwork also showed a significant relation with follow-up activities. The influence of non-measured variables in the model was considerable. CONCLUSION: In exploring the organisational setting as a barrier to prevention and disease management, the designed model showed no major effects. Despite the wide variety of practice organisational items investigated, a strong influence of non-measured variables was evident. Teamwork in the practices proved to be related to both follow-up and record-keeping. Younger and female GPs were further predictors of adequate record-keeping.

Process evaluation of a multifaceted intervention to improve cardiovascular disease prevention in general practice.

OBJECTIVES: To perform a process evaluation of a multifaceted intervention to improve cardiovascular and diabetes care in general practice. METHODS: The feasibility of the intervention, carried out by outreach visitors in 62 practices, was addressed by evaluating whether the intervention programme was performed as planned and the extent to which it was accepted by the practice team. In addition, the costs of the programme were determined. RESULTS: The intervention was largely carried out as planned, although the intervention period had to be extended by three months. Of the 18 topics that could be addressed during the intervention period, 12 (mean) were addressed. The number of outreach visits per practice was 15.2 (mean), each visit lasted about one hour. Most practice members endorsed both the key recommendations for clinical decision-making and cardiovascular risk profiling. The majority of GPs (range 63-98%) agreed with the guidelines for clinical decision-making, and 29-97% had a positive opinion about the guidelines for practice organisation. According to practice staff members, the outreach visitor had sufficient knowledge and skills to support them in changing the practice organisation. GPs were less positive about the outreach visitor's knowledge and skills in optimising clinical decision-making; however 78% believed that the outreach visitor contributed to effecting change in their clinical decision-making. The total costs of the intervention per practice were Euro 4317. CONCLUSIONS: This process evaluation demonstrated that the intervention was usually carried out as planned and achieved a high satisfaction rating from the participating practice members.

Deprivation and systematic stroke prevention in general practice: an audit among general practitioners in the Rotterdam region, The Netherlands.

BACKGROUND: To investigate differences in quality of preventive care provided by general practitioners (GPs) to patients at risk of stroke living in deprived and non-deprived neighbourhoods in the Rotterdam region. METHODS: A 'deprivation score' was used to categorize neighbourhoods according to their deprivation status. Data on the process of patient care were collected by means of chart review and interviews with GPs. Cases of stroke (n=188) were retrospectively audited by an expert panel with guideline-based review criteria. To measure differences in quality of patient care between neighbourhoods, deprivation scores were related to scores for sub-optimal care. RESULTS: After adjustment for socio-demographic characteristics, patients in deprived neighbourhoods had an increased risk (OR 1.95 (95% CI: 0.98-3.90)) of having received sub-optimal preventive care if compared with patients in non-deprived neighbourhoods. This excess risk was limited to women (OR 3.57 (95% CI: 1.39-9.16) vs OR 1.01 (95% CI: 0.41-2.48) in men). Adjustment for socio-demographic characteristics and risk factor distribution did not change the OR for women to receive sub-optimal care significantly (OR 3.21 (95% CI: 1.24-8.31)). Sub-optimal care originated mainly from deficiencies in follow-up of treated hypertensive and diabetes patients and evaluation of patients' cardiovascular risk profile. Among treated hypertensive women in deprived neighbourhoods who received sub-optimal care, the mean number of deficiencies related to follow-up was almost double that of the corresponding group in non-deprived neighbourhoods. CONCLUSION: Quality of care to prevent stroke in general practice differs considerably between deprived and non-deprived neighbourhoods. Patients in deprived neighbourhoods, and women in particular, have almost twice the risk of receiving sub-optimal preventive care.

Improving quality of organizing cardiovascular preventive care in general practice by outreach visitors: a randomized controlled trial.

BACKGROUND: Adequate care for patients with cardiovascular risks requires an adequate practice organization. Educational outreach visits are a promising approach to modifying professional behavior. We aimed to assess whether the quality of cardiovascular preventive care in general practice can be improved through a comprehensive intervention implemented by an educational outreach visitor. METHODS: After baseline measurements, general practices (n = 124) in the southern half of The Netherlands were randomly allocated to either intervention or control group. The intervention, based on the educational outreach model, comprised 15 practice visits over a period of 21 months and addressed a large number of issues around task delegation, availability of instruments and patient leaflets, record-keeping, and follow-up routines. Twenty-one months after the start of the intervention, postintervention measurements were performed. The difference between ideal and actual practice in each aspect of organizing preventive care was defined as a deficiency score. Primary outcome measure was the difference in deficiency scores before and after the intervention. RESULTS: All practices completed both baseline and postintervention measurements. The difference in change between intervention and control group adjusted for baseline was statistically significant (P < 0.001) for each aspect of organizing preventive care. The largest absolute improvement was found for the number of preventive tasks performed by the practice assistant. CONCLUSIONS: This study showed that a comprehensive intervention implemented by outreach visitors was effective in improving organization of cardiovascular preventive care in general practice.

Organizing cardiovascular preventive care in general practice: determinants of a successful intervention.

BACKGROUND: Although outreach visitor interventions have proven to be effective, more detailed studies are needed to understand what elements of interventions work and why. In this study we investigate the determinants of success of an intervention for optimizing cardiovascular preventive care in general practice. METHODS: After baseline measurements and randomization, 62 general practices received a comprehensive intervention program, by means of outreach visitors, lasting 21 months. Data on practice management and preventive activities were gathered at baseline and at postintervention measurements. Key characteristics of the intervention considered possible determinants of success were gathered by questionnaire. The difference between ideal and actual practice in each aspect of organizing cardiovascular preventive care was calculated as a deficiency score. The difference between deficiency scores before and after the intervention were the main outcome measures. RESULTS: The key characteristic, duration of exposure to an aspect (in months), was positively related to the change in availability of separate clinics and in the amount of teamwork. The improvement in instruments and materials was positively related to the general practitioner's opinion about the given feedback. No relations were found between the key characteristics and changes in record-keeping or follow-up routines. CONCLUSIONS: Although implementation of a comprehensive prevention program is effective, we could not fully disentangle the "black box" of the intervention. The duration of exposure to an aspect of organizing cardiovascular care was the key determinant to success.

Provision of information and advice in cardiovascular care: clinical performance of general practitioners.

The aim of this study was to assess the provision of information and advice by general practitioners (GPs) with respect to patients with hypertension, hypercholesterolaemia, or cardiovascular disease. The study relied on the prospective recording of patient encounters by GPs. Performance indicators were selected from the Dutch national guidelines for general practice. The GPs (n=195) completed 5330 encounter forms. High levels of performance were found with regard to advice on smoking cessation and the provision of information (e.g. information about alarm symptoms or the aim of treatment). Low levels of performance were found with regard to advice on salt consumption, alcohol consumption, weight reduction and physical exercise for patients with hypertension. Discussion of compliance with the therapy in case of hypercholesterolaemia, advice on physical exercise in case of angina pectoris and advice on foot care in case of peripheral arterial disease also showed a substantial gap between recommended and actual care. Performance rates varied considerably across GPs. The patient and GP characteristics examined in this study contributed very little to the clinical performance.

Normal values and determinants of circadian urine production in older men: a population based study.

PURPOSE: We evaluated circadian urine production and its determinants in a large population based sample of older men. MATERIALS AND METHODS: We collected data on 1,688 men 50 to 78 years old, without radical prostatectomy, prostate or bladder cancer, neurogenic bladder disease or negative advice from their general practitioner, recruited from the population of Krimpen, the Netherlands. Measurements consisted of self-administered questionnaires, including the International Prostate Symptom Score, a 3-day frequency-volume chart, transrectal prostatic ultrasound, uroflowmetry and post-void residual volume. Hourly urine production was determined and urine production day-to-night ratio was calculated from the frequency-volume chart. RESULTS: Men younger than 65 years showed a clear circadian urine production pattern, whereas in older men this pattern was less clear. Smoking, use of diuretic drugs, post-void residual and 24-hour polyuria reinforced the circadian pattern, in favor of daytime urine production. The urine production day-to-night ratio was not associated with prostate enlargement, reduced urinary flow rate, body weight, hypertension, cardiac symptoms, diabetes mellitus, use of antidepressants, cardiac or hypnotic drugs. CONCLUSIONS: Urine production in men younger than 65 years showed a clear circadian pattern in contrast to men older than 65 years. These data can be used as a reference when describing urine production patterns in select populations. In daily practice frequency-volume charts can be used to determine urine production. This method is inexpensive, easy to use and provides valid information on urine production in a natural environment.

Health status and its correlates among Dutch community-dwelling older men with and without lower urogenital tract dysfunction.

OBJECTIVE: To study health status and its correlates in older men with and without lower urogenital tract dysfunction. METHODS: Cross-sectional population-based study on 1688 men aged 50-78 years without bladder or prostate cancer, radical prostatectomy, neurogenic bladder dysfunction or a negative advice from their general practitioner. Data were collected through self-administered questionnaires, including Sickness Impact Profile (SIP, three domains), Inventory of Subjective Health (ISH), International Prostate Symptom Score (IPSS) and International Continence Society (ICS) Male Sex questionnaire, medication use, socio-economic and lifestyle factors. Additional information was collected by measurement of blood pressure, transrectal ultrasonography of the prostate and uroflowmetry. Four health status domains were analyzed using the ISH and three domains of the SIP. Lower urinary tract symptoms (LUTS) were categorised using IPSS, erectile and ejaculatory dysfunction were defined using the ICS questionnaire. RESULTS: All urogenital characteristics and parameters were related to at least two of the health status domains. Multivariate regression analyses yielded that LUTS and cardiac symptoms were associated with suboptimal scores of all four domains. Chronic obstructive pulmonary disease and drugs for abdominal symptoms were related to three domains; erectile and ejaculatory dysfunction, muskuloskeletal or psycho(ana)leptic drugs and marital status to two domains. CONCLUSIONS: The impact of LUTS on health status was equally important as the impact of cardiac symptoms. The impact of sexual dysfunction was smaller than expected. Longitudinal studies are needed to determine how health status and illnesses interact.