Evidence-based Healthcare

Rate of major complications is higher in laparoscopic than abdominal hysterectomy but quality of life improves with both procedures

Lee A Learman — 2004-08-01

Abstract: Question: Is the risk of major complications greater following laparoscopic hysterectomy compared with abdominal and vaginal hysterectomy for non-malignant conditions? Study design: Two parallel, multi-centre randomised trials. Main results: More major complications were experienced with laparoscopic hysterectomy compared with abdominal hysterectomy (11.1% vs 6.2%; mean difference 4.9%, 95% CI 0.9% to 9.1%, number needed to harm 20). There was no significant difference in complication rates between laparoscopic and vaginal hysterectomy groups (complication rate 9.5% for both groups). Pain scores were higher following abdominal hysterectomy compared with laparoscopic hysterectomy (mean difference 0.4, 95% CI 0.09 to 0.7). There was no detectable difference in the vaginal trial. Quality of life at 12 months improved with all interventions. Authors’ conclusions: Major complications were more common following laparoscopic hysterectomy compared with abdominal hysterectomy. The vaginal trial was inconclusive.

Consuming less than 4 alcoholic drinks per week does not increase risk of pre-term delivery

Heather L. Paladine — 2004-08-01

Abstract: Question: Is there an association between the amount and type of alcohol consumed during pregnancy and the risk of pre-term delivery? Study design: Prospective cohort study. Main results: Of 40, 892 pregnant women, 1880 (4.6%) had pre-term delivery. The adjusted relative risk of pre-term delivery in all women consuming 2–3.5 drinks per week was lower than in non-drinkers (RR 0.80, 95% CI 0.68 to 0.96). The risk was not statistically significant when only nulliparous women were included. Other levels of alcohol consumption were not associated with a statistically significant increased or decreased risk of pre-term delivery compared with non-drinkers except for nulliparous women who drank ⩾7 drinks per week (RR 2.91, 95% CI 1.29 to 6.55). There was no relationship between risk of pre-term delivery and the preferred type of alcohol (wine, beer, spirits or mixed). Authors’ conclusions: Consumption of 7 or more drinks per week was associated with an increased risk of pre-term delivery in woman having their first child. Pre-term delivery was not affected by type of alcohol intake.

Appointment attendance predicts level of glycaemic control in people with diabetes

Barbara M B.M. Rohland — 2004-08-01

Abstract: Question: Do people who miss scheduled healthcare appointments have poorer management of their diabetes? Study design: Retrospective cross-sectional cohort study. Main results: During the year 2000, 12% of participants missed >30% of scheduled appointments, 52% missed 1–30% and 35% missed none. Glycaemic control was poorer in those who missed most appointments (p<0.0001), irrespective of how diabetes was controlled (diet, oral agent or insulin). The adjusted mean glycosylated haemoglobin level (HbA1c) was higher in people attending appointments compared with people missing >30% of appointments (p<0.0001). Daily self-monitoring was associated with a fewer missed appointments (OR 1.8, 95% CI 1.7 to 1.9). People with inadequate medication ⩾20% of the time were more likely to miss appointments (OR 1.5, p<0.0001). Living in a poverty area, pharmacologic control of diabetes, use of antidepressants and having fewer scheduled appointments were significantly associated with missing >30% of scheduled appointments. Authors’ conclusions: People with diabetes who frequently miss appointments are likely to have poorer glycaemic control and less frequent self-monitoring. Level of attendance could be monitored and used for clinical risk stratification.

Unscheduled care for people with asthma in a multi-ethnic area is reduced following educational outreach programme by specialist nurses

Arvid W.A A.W.A. Kamps — 2004-08-01

Abstract: Question: Do asthma specialist nurses reduce health service use and improve outcomes across different ethnic groups? Study design: Cluster randomised controlled trial. Main results: Specialist nurse intervention reduced the percentage of participants attending unscheduled care in the subsequent year compared with usual care (adjusted odds ratio with clustering 0.61, 95% CI 0.38 to 0.99). First attendance for unscheduled asthma care in the same time period was delayed by specialist nurse intervention (hazard ratio [HR] 0.73, 95% CI 0.54 to 1.00). The effect of specialist nurse intervention on time to attendance was greater in white people (HR 0.57, 95% CI 0.38 to 0.85) compared with South Asian people (HR 0.72, 95% CI 0.48 to 1.09) or other ethnicities (HR 1.29, 95% CI 0.51 to 3.22). There were no significant difference in mean rates of hospital admission between groups. Authors’ conclusions: Unscheduled care for people with asthma was reduced in practices where asthma specialist nurses provided an educational outreach and clinical support programme to staff. Improved outcomes were not equally distributed among ethnic groups.

Liver cancer screening in a high-risk population in China fails to reduce mortality

Sammy Saab — 2004-08-01

Abstract: Question: Does liver cancer screening reduce mortality from the disease in a high-risk Chinese population? Study design: Cluster randomised controlled trial. Main results: In people at high risk of liver cancer, screening did not significantly reduce the incidence of primary liver cancer or risk of death compared with no screening (see Table 1), despite earlier detection of the disease (see notes). †Per 100,000 person years. Authors’ conclusions: Liver cancer screening in a high-risk population in China does not reduce mortality from the disease.

Breast feeding support from volunteer counsellors does not increase rates of breast feeding

Amal K Mitra — 2004-08-01

Abstract: Question: Does breast feeding support from volunteer counsellors increase the number of women who breast feed? Study design: Randomised controlled trial. Main results: At both 6 weeks and 4 months, there was no significant difference in rates of breast feeding between women receiving breast feeding counselling and those that did not (see Table 1). Authors’ conclusions: Breast feeding support by volunteer counsellors did not increase breast feeding rates.

Lowering homocysteine levels does not reduce rates of stroke, coronary heart disease or death in people with ischaemic stroke

Yi-Chia Y.-C. Huang — 2004-08-01

Abstract: Question: Do high doses of vitamins B6 and B12 and folic acid, which lower total homocysteine levels, reduce the risk of recurrent stroke, myocardial infarction or death compared with low doses in people with non disabling acute ischaemic stroke? Study design: Multicentre randomised controlled trial. Main results: High dose vitamins reduced mean total homocysteine levels by 2 μmol/l more than low dose vitamins. In people surviving a non disabling acute ischaemic stroke, no significant differences in rates of recurrent ischaemic stroke, coronary heart disease or death occurred between people receiving high dose or low dose vitamins over 2 years (see Table 1). Authors’ conclusions: In people with ischaemic stroke, lowering homocysteine levels does not reduce the rate of recurrent stroke, coronary heart disease or death.

Smoking high tar cigarettes increases risk of death from lung cancer, but no differences in risk for smokers of very low, low and medium tar cigarettes

Silvano Gallus — 2004-08-01

Abstract: Question: What is the relationship between lung cancer mortality and tar rating of smoker's cigarette brand? Study Design: Multivariate analysis of data from prospective cohort study. Main results: Risk of death due to lung cancer was significantly higher in smokers of high-tar cigarettes compared with people smoking low, very low and medium tar cigarettes (see Table 1). There were no significant differences in the risk of lung cancer mortality for people smoking very low or low tar cigarettes compared with smokers of medium tar cigarettes. Stopping smoking considerably reduced risk of lung cancer. People quitting before age 35 years had a very similar risk to those who had never smoked. Authors’ conclusions: There was no detectable difference in risk of lung cancer among people who smoked very low, low or medium tar cigarettes. An increased risk was identified in people who smoked high tar cigarettes. These findings persisted after adjustment for demographics, diet and medical history and for cigarettes/day.

H. pylori eradication does not reduce gastric cancer incidence in a high-risk area of China

Bor-Shyang Sheu, Xi-Zhang Lin — 2004-08-01

Abstract: Question: Does treating H. Pylori reduce the incidence of gastric cancer? Study: Randomised placebo-controlled trial. Main results: At 7.5 years, incidence of gastric cancer was not significantly different between groups (treatment: 0.86%, n=7; placebo: 1.35%, n=11; p=0.33). In people without precancerous lesions at baseline, risk of gastric cancer was significantly lower for treatment compared with placebo (incidence of gastric cancer: treatment n=0; placebo n=6; p=0.02) but there was no difference among people with precancerous lesions (n=7 and 5, respectively). Smoking (hazard ratio [HR] 6.2, 95% CI 2.3 to 16.5) and older age (HR per 1 year 1.10, 95% CI 1.05 to 1.15) were independent predictors of gastric cancer risk. Authors’ conclusions: Incidence of gastric cancer was similar between H. pylori eradication and placebo groups over 7.5 years of follow-up. However, it was significantly reduced in a subgroup of people without precancerous lesions at baseline.

Hormone replacement therapy is not safe for breast cancer survivors

Rena Vassilopoulou-Sellin — 2004-08-01

Abstract: Question: Is hormone replacement therapy safe for women with previous breast cancer? Study design: Randomised controlled trial (interim analysis). Main results: In 345 women surviving breast cancer, there were more new breast cancer events in women taking HRT for menopausal symptoms compared with women receiving symptomatic treatment without hormones at a median of 2 years follow-up (absolute risk for new breast cancer: 26/174 [14%] with HRT vs 8/171 [5%] with no HRT; relative hazard 3.5, 95% CI 1.5 to 8.1). Authors’ conclusions: In women surving breast cancer, those who received HRT for menopausal symptoms were at a higher risk of developing new breast cancers compared with those who received symptomatic treatment without hormones. These findings led to the termination of the trial.

Practice-based continuing education combined with process improvement methods improves delivery of preventive services to children

James L. J. Vacek — 2004-08-01

Abstract: Question: Does continuing medical education in combination with process improvement methods to implement office systems, increase rates of delivery of preventive care to children? Study design: Cluster randomised controlled trial. Main results: Significantly more children enrolled in intervention practices received all four preventive services compared with control practices after 30 months (change in proportion of children receiving all four preventive services: 7% to 34% with intervention vs 9% to 10% with no intervention; 4.6-fold increase with intervention compared with control, 95% CI 1.6 to 13.2, see Table 1). Authors’ conclusions: Practice-based continuing medical education in combination with process improvement methods increases the rate of delivery of preventive care services to children.

Retaplase plus abciximab improves non-fatal outcomes, but not overall survival in people with diabetes and acute ST-segment elevation myocardial infarction

Victor Serebruany — 2004-08-01

Abstract: Question: In people with diabetes and ST-segment elevation myocardial infarction, is half-dose reteplase plus abciximab more effective than reteplase alone? Study design: Secondary analysis of multicentre randomised controlled trial. Main results: For mortality at 30 days or 1 year, there was no significant difference between groups (see Table 1). Reteplase plus abciximab significantly reduced the risk of reinfarction, or recurrent ischaemia or angina compared with reteplase alone. Authors’ conclusions: Although treatment with reteplase plus abciximab did not provide a survival benefit for people with diabetes and ST-segment elevation myocardial infarction compared with reteplase alone, nonfatal outcomes including reinfarction and recurrent ischaemia were substantially reduced.

No association between mobile phone usage and development of acoustic neuroma

Lennart Hardell — 2004-08-01

Abstract: Question: Is the incidence of acoustic neuroma associated with mobile phone usage? Study design: Population-based case-control study. Main results: Risk of acoustic neuroma was not statistically higher among people regularly using mobile phones compared with those who never or rarely using one (OR 0.90, 95% CI 0.51 to 1.57). No increased risk was associated with the length of time since beginning regular use of mobile phones, total duration or number of calls (see Table 1). In people with acoustic neuroma, mean tumour size was bigger in regular mobile phone users (1.66cm3) than non-users (1.39cm3) (Wilcoxon, p=0.03) but the increased risk of developing a large tumour (⩾1.51cm3) was not statistically significant (OR 1.87, 95% CI 0.75 to 4.64). Authors’ conclusions: Mobile phone usage is not positively associated with the development of acoustic neuroma, regardless of the estimated total duration or number of calls, or the length of time since regular use began.

Link between diagnostic X-rays and cancer uncertain

John R Cameron — 2004-08-01

Abstract: Question: Do diagnostic X-rays increase the risk of developing cancer? Study design: Epidemiological risk analysis. Main results: By age 75, the cumulative risks of cancer resulting from diagnostic X-rays were estimated to be between 0.6% (in the UK and Poland) and 3.2% (Japan). In all other populations the estimated risk was between 0.7% and 1.8%. The UK risk correlates to 700 cases per year including 111 bladder cancers, 107 colon cancers and 61 lung cancers. Authors’ conclusions: According to their model, the estimated cumulative risk of cancer caused by diagnostic X-rays varied between countries, however most were fewer than 2%. In Japan the estimated cumulative risk exceeded 3%.

Chest pain observation units reduce hospital admission in people with acute chest pain

Michael A Ross — 2004-08-01

Abstract: Question: For people presenting with acute chest pain, how effective and cost-effective is care in a chest pain observation unit compared with routine care? Study design: Cluster randomised controlled trial. Main results: Hospital admissions were significantly less likely with the chest pain observation unit (CPOU) compared with routine care (36.7% CPOU vs 53.8% routine care; adjusted OR 0.49, 95% CI 0.36 to 0.65). There was no significant difference between groups in the incidence of major coronary events at 6 months (3.8% with CPOU vs 3.4% with routine care; difference 0.4%, 95% CI –2.0% to +2.7%). The mean cost of chest pain related care per participant was lower for the CPOU than for routine care, but the difference was not significant (£478 for CPOU vs £556 for routine care; adjusted difference £53, 95% CI −£88 to +£194). Authors’ conclusions: Care in a chest pain observation unit reduces hospital admissions and may be more cost effective than routine care.

Exercise training programmes improve survival and delay hospital admission in people with chronic heart failure

Neil Smart, Thomas H T. Marwick — 2004-08-01

Abstract: Question: Do exercise training programmes improve survival in people with chronic heart failure? Study design: Systematic review with meta-analysis. Main results: Fewer deaths occurred in the exercise group compared with control at a median follow-up of about 2 years (exercise: 22% [88/395]; control 26% [105/406]; hazard ratio 0.65, 95% CI 0.46 to 0.92). Fewer deaths or hospital admissions occurred in the exercise group compared with control (exercise: 32% [127/395]; control 43% [173/406]; hazard ratio HR 0.72, 95% CI 0.56 to 0.93). Authors’ conclusions: Exercise training improves survival and time to death or admission to hospital in people with chronic heart failure due to left ventricular systolic dysfunction.

Laparoscopic hysterectomy is not cost effective compared with vaginal hysterectomy

Benjamin L Crawford — 2004-08-01

Abstract: Question: What is the cost effectiveness of laparoscopic hysterectomy compared with conventional (vaginal or abdominal) hysterectomy? Study design: Two parallel multicentre randomised controlled trials. Main results Vaginal hysterectomy:: Laparoscopic hysterectomy was significantly more expensive than vaginal hysterectomy, however there were no differences in quality adjusted life years (QALYs; see Table 1). Abdominal hysterectomy:: There were no significant differences between abdominal and laparoscopic hysterectomy for cost or QALYs. For each additional QALY, laparoscopic hysterectomy was estimated to cost £267,333 more than vaginal hysterectomy, and £26,571 more than abdominal hysterectomy. QALY, quality adjusted life year; ICER, incremental cost effectiveness ratio (mean difference in cost divided by mean difference in QALYs) Authors’ conclusions: Laparoscopic hysterectomy does not offer any cost-effectiveness benefit over vaginal hysterectomy. Laparoscopic hysterectomy was similarly cost effective to abdominal hysterectomy.

Intravenous magnesium sulphate does not improve survival or disability outcomes in people with stroke

Scott Selco, Bruce Ovbiagele — 2004-08-01

Abstract: Question: Does intravenous magnesium sulphate improve survival or disability outcomes if given within 12 hours of stroke onset? Study design: Multicentre randomised controlled trial. Main results: Intravenous magnesium sulphate did not significantly improve survival or disability outcomes at 90 days compared with placebo in people with stroke (see Table 1). Authors’ conclusions: Intravenous magnesium sulphate given within 12 hours of stroke does not improve survival or disability outcomes.

Attending a single care site associated with improved glycaemic control in people with diabetes

Peter Harvey — 2004-08-01

Abstract: Question: Does continuity of care improve control of clinical risk factors in people with diabetes? Study design: Cross-sectional community-based survey. Main results: 85.5% of participants had continuity of care (single care site and usual provider), 9.3% had a single care site but different providers, and 5.2% had no usual source of care. Good glycaemic control was more likely with continuity of care or single care site compared with no usual source of care (continuity of care: OR 4.62, 95% CI 2.02 to 10.60; single care site: OR 6.13, 95% CI 2.08 to 18.04). There were no significant differences between groups with a usual site. There was no increased likelihood of good control of blood pressure or lipid level among groups. Authors’ conclusions: There is evidence that good glycaemic control is more likely among people whose diabetic care is provided from one site, regardless of whether it is provided by the same practitioner.

After a stroke, ability with daily tasks of living improves after therapy based rehabilitation services

Michael Power — 2004-08-01

Abstract: Question: Do post-discharge rehabilitation services change recovery after stroke? Study design: Systematic review with meta-analysis. Main results: Fourteen trials met inclusion criteria; 12 trials comparing therapy-based rehabilitation services were included in the meta-analysis (occupational therapy = 6 trials, physiotherapy = 2 trials and mixed services = 4 trials). At a median follow-up of 6 months, therapy-based rehabilitation services reduced the risk of deterioration in ability to undertake daily living tasks compared with control, (OR 0.72, 95% CI 0.57 to 0.92). Ability to carry out extended activities of daily living significantly improved in people undergoing therapy-based rehabilitation services compared with control (mean difference 0.17 95% CI 0.04 to 0.30). When similar categories of therapy were compared, only occupational therapy significantly reduced deterioration rate (occupational therapy: OR 0.73, 95% CI 0.55 to 0.96; physiotherapy OR 0.67, 95% CI 0.24 to 1.89; mixed services OR 0.72, 95% CI 0.41 to 1.27). Data were inconclusive with respect to mood, quality of life, need for long-term care and hospital readmission. Authors’ conclusions: People discharged to their homes after stroke are less likely to deteriorate if therapy-based rehabilitation services are provided compared with usual care or no routine intervention.

Intensive follow-up for colorectal cancer is cost-effective

Björn Ohlsson — 2004-08-01

Abstract: Question: Is intensive follow-up after curative resection for colorectal cancer cost-effective compared with conventional follow-up methods? Study design: Incremental cost-effectiveness analysis based on a meta-analysis of 5 RCTs. Main results: Over 5 years, people having intensive follow-up gained 0.73 to 0.82 life years compared with conventional follow-up (absolute reduction in mortality with intensive follow-up: all trials 7%; 95% CI 5% to 9%; extramural trials only 9%; 95% CI 7% to 11%). Additional costs incurred by intensive follow-up were £3402 per life year gained (or £3077 based on extramural trials only). The net costs per person were £2479 and £2529, respectively. Authors’ conclusions: Intensive follow-up of people with colorectal cancer is cost-effective compared to conventional methods, based on a robust economic model and efficacy evidence from 5 RCTs.

Six monthly follow-up is as effective as 3 monthly follow-up for people with hypertension

Mark R Nelson — 2004-08-01

Abstract: Question: In people with treated hypertension, how effective is 3 month follow-up by a family physician compared with 6 month follow-up, in terms of blood pressure control, patient satisfaction and adherence to drug treatment over 3 years? Study design: Randomised controlled trial. Main results: At 36 months, mean blood pressure, patient satisfaction and adherence to treatment was equivalent between treatment groups (see Table 1). Authors’ conclusions: In people with controlled hypertension, 6 monthly follow-up is as effective as 3 monthly follow-up over 3 years for control of blood pressure, patient satisfaction and adherence to treatment, and has no effect on the proportion of participants with out of control blood pressure.