Comparative Effectiveness of Dynamic Treatment Strategies for Medication Use and Dosage: Emulating a Target Trial Using Observational Data

BACKGROUND

Availability of detailed data from electronic health records (EHRs) has increased the potential to examine the comparative effectiveness of dynamic treatment strategies using observational data. Inverse probability (IP) weighting of dynamic marginal structural models can control for time-varying confounders. However, IP weights for continuous treatments may be sensitive to model choice.

METHODS

We describe a target trial comparing strategies for treating anemia with darbepoetin in hemodialysis patients using EHR data from the UK Renal Registry 2004 to 2016. Patients received a specified dose (microgram/week) or did not receive darbepoetin. We compared 4 methods for modeling time-varying treatment: (A) logistic regression for zero dose, standard linear regression for log dose; (B) logistic regression for zero dose, heteroscedastic linear regression for log dose; (C) logistic regression for zero dose, heteroscedastic linear regression for log dose, multinomial regression for patients who recently received very low or high doses; and (D) ordinal logistic regression.

RESULTS

For this dataset, method (C) was the only approach that provided a robust estimate of the mortality hazard ratio (HR), with less-extreme weights in a fully weighted analysis and no substantial change of the HR point estimate after weight truncation. After truncating IP weights at the 95th percentile, estimates were similar across the methods.

CONCLUSIONS

EHR data can be used to emulate target trials estimating the comparative effectiveness of dynamic strategies adjusting treatment to evolving patient characteristics. However, model checking, monitoring of large weights, and adaptation of model strategies to account for these is essential if an aspect of treatment is continuous.

Natural course of chronic fatigue syndrome/myalgic encephalomyelitis in adolescents

OBJECTIVE

Little is known about persistence of or recovery from chronic fatigue syndrome/myalgic encephalomyelitis (CFS/ME) in adolescents. Previous studies have small sample sizes, short follow-up or have focused on fatigue rather than CFS/ME or, equivalently, chronic fatigue, which is disabling. This work aimed to describe the epidemiology and natural course of CFS/ME in adolescents aged 13-18 years.

DESIGN

Longitudinal follow-up of adolescents enrolled in the Avon Longitudinal Study of Parents and Children.

SETTING

Avon, UK.

PARTICIPANTS

We identified adolescents who had disabling fatigue of >6 months duration without a known cause at ages 13, 16 and 18 years. We use the term 'chronic disabling fatigue' (CDF) because CFS/ME was not verified by clinical diagnosis. We used multiple imputation to obtain unbiased estimates of prevalence and persistence.

RESULTS

The estimated prevalence of CDF was 1.47% (95% CI 1.05% to 1.89%) at age 13, 2.22% (1.67% to 2.78%) at age 16 and 2.99% (2.24% to 3.75%) at age 18. Among adolescents with CDF of 6 months duration at 13 years 75.3% (64.0% to 86.6%) were not classified as such at age 16. Similar change was observed between 16 and 18 years (75.0% (62.8% to 87.2%)). Of those with CDF at age 13, 8.02% (0.61% to 15.4%) presented with CDF throughout the duration of adolescence.

CONCLUSIONS

The prevalence of CDF lasting 6 months or longer (a proxy for clinically diagnosed CFS/ME) increases from 13 to 18 years. However, persistent CDF is rare in adolescents, with approximately 75% recovering after 2-3 years.

Screening strategies for atrial fibrillation: a systematic review and cost-effectiveness analysis

Background

Atrial fibrillation (AF) is a common cardiac arrhythmia that increases the risk of thromboembolic events. Anticoagulation therapy to prevent AF-related stroke has been shown to be cost-effective. A national screening programme for AF may prevent AF-related events, but would involve a substantial investment of NHS resources.

Objectives

To conduct a systematic review of the diagnostic test accuracy (DTA) of screening tests for AF, update a systematic review of comparative studies evaluating screening strategies for AF, develop an economic model to compare the cost-effectiveness of different screening strategies and review observational studies of AF screening to provide inputs to the model.

Design

Systematic review, meta-analysis and cost-effectiveness analysis.

Setting

Primary care.

Participants

Adults.

Intervention

Screening strategies, defined by screening test, age at initial and final screens, screening interval and format of screening {systematic opportunistic screening [individuals offered screening if they consult with their general practitioner (GP)] or systematic population screening (when all eligible individuals are invited to screening)}.

Main outcome measures

Sensitivity, specificity and diagnostic odds ratios; the odds ratio of detecting new AF cases compared with no screening; and the mean incremental net benefit compared with no screening.

Review methods

Two reviewers screened the search results, extracted data and assessed the risk of bias. A DTA meta-analysis was perfomed, and a decision tree and Markov model was used to evaluate the cost-effectiveness of the screening strategies.

Results

Diagnostic test accuracy depended on the screening test and how it was interpreted. In general, the screening tests identified in our review had high sensitivity (> 0.9). Systematic population and systematic opportunistic screening strategies were found to be similarly effective, with an estimated 170 individuals needed to be screened to detect one additional AF case compared with no screening. Systematic opportunistic screening was more likely to be cost-effective than systematic population screening, as long as the uptake of opportunistic screening observed in randomised controlled trials translates to practice. Modified blood pressure monitors, photoplethysmography or nurse pulse palpation were more likely to be cost-effective than other screening tests. A screening strategy with an initial screening age of 65 years and repeated screens every 5 years until age 80 years was likely to be cost-effective, provided that compliance with treatment does not decline with increasing age.

Conclusions

A national screening programme for AF is likely to represent a cost-effective use of resources. Systematic opportunistic screening is more likely to be cost-effective than systematic population screening. Nurse pulse palpation or modified blood pressure monitors would be appropriate screening tests, with confirmation by diagnostic 12-lead electrocardiography interpreted by a trained GP, with referral to a specialist in the case of an unclear diagnosis. Implementation strategies to operationalise uptake of systematic opportunistic screening in primary care should accompany any screening recommendations.

Limitations

Many inputs for the economic model relied on a single trial [the Screening for Atrial Fibrillation in the Elderly (SAFE) study] and DTA results were based on a few studies at high risk of bias/of low applicability.

Future work

Comparative studies measuring long-term outcomes of screening strategies and DTA studies for new, emerging technologies and to replicate the results for photoplethysmography and GP interpretation of 12-lead electrocardiography in a screening population.

Study registration

This study is registered as PROSPERO CRD42014013739.

Funding

The National Institute for Health Research Health Technology Assessment programme.

Using clinical practice variations as a method for commissioners and clinicians to identify and prioritise opportunities for disinvestment in health care: a cross-sectional study, systematic reviews and qualitative study

Background

NHS expenditure has stagnated since the economic crisis of 2007, resulting in financial pressures. One response is for policy-makers to regulate use of existing health-care technologies and disinvest from inefficiently used health technologies. A key challenge to disinvestment is to identify existing health technologies with uncertain cost-effectiveness.

Objectives

We aimed to explore if geographical variation in procedure rates is a marker of clinical uncertainty and might be used by local commissioners to identify procedures that are potential candidates for disinvestment. We also explore obstacles and solutions to local commissioners achieving disinvestment, and patient and clinician perspectives on regulating access to procedures.

Methods

We used Hospital Episode Statistics to measure geographical variation in procedure rates from 2007/8 to 2011/12. Expected procedure numbers for each primary care trust (PCT) were calculated adjusting for proxies of need. Random effects Poisson regression quantified the residual inter-PCT procedure rate variability. We benchmarked local procedure rates in two PCTs against national rates. We conducted rapid systematic reviews of two high-use procedures selected by the PCTs [carpal tunnel release (CTR) and laser capsulotomy], searching bibliographical databases to identify systematic reviews and randomised controlled trials (RCTs). We conducted non-participant overt observations of commissioning meetings and semistructured interviews with stakeholders about disinvestment in general and with clinicians and patients about one disinvestment case study. Transcripts were analysed thematically using constant comparison methods derived from grounded theory.

Results

There was large inter-PCT variability in procedure rates for many common NHS procedures. Variation in procedure rates was highest where the diffusion or discontinuance was rapidly evolving and where substitute procedures were available, suggesting that variation is a proxy for clinical uncertainty about appropriate use. In both PCTs we identified procedures where high local use might represent an opportunity for disinvestment. However, there were barriers to achieving disinvestment in both procedure case studies. RCTs comparing CTR with conservative care indicated that surgery was clinically effective and cost-effective on average but provided limited evidence on patient subgroups to inform commissioning criteria and achieve savings. We found no RCTs of laser capsulotomy. The apparently high rate of capsulotomy was probably due to the coding inaccuracy; some savings might be achieved by greater use of outpatient procedures. Commissioning meetings were dominated by new funding requests. Benchmarking did not appear to be routinely carried out because of capacity issues and concerns about data reliability. Perceived barriers to disinvestment included lack of collaboration, central support and tools for disinvestment. Clinicians felt threshold criteria had little impact on their practice and that prior approval systems would not be cost-effective. Most patients were unaware of rationing.

Conclusions

Policy-makers could use geographical variation as a starting point to identify procedures where health technology reassessment or RCTs might be needed to inform policy. Commissioners can use benchmarking to identify procedures with high local use, possibly indicating overtreatment. However, coding inconsistency and limited evidence are major barriers to achieving disinvestment through benchmarking. Increased central support for commissioners to tackle disinvestment is needed, including tools, accurate data and relevant evidence. Early engagement with patients and clinicians is essential for successful local disinvestment.

Funding

The National Institute for Health Research Health Services and Delivery Research programme.

Evaluation of the Cochrane Collaboration's tool for assessing the risk of bias in randomized trials: focus groups, online survey, proposed recommendations and their implementation

BACKGROUND

In 2008, the Cochrane Collaboration introduced a tool for assessing the risk of bias in clinical trials included in Cochrane reviews. The risk of bias (RoB) tool is based on narrative descriptions of evidence-based methodological features known to increase the risk of bias in trials.

METHODS

To assess the usability of this tool, we conducted an evaluation by means of focus groups, online surveys and a face-to-face meeting. We obtained feedback from a range of stakeholders within The Cochrane Collaboration regarding their experiences with, and perceptions of, the RoB tool and associated guidance materials. We then assessed this feedback in a face-to-face meeting of experts and stakeholders and made recommendations for improvements and further developments of the RoB tool.

RESULTS

The survey attracted 380 responses. Respondents reported taking an average of between 10 and 60 minutes per study to complete their RoB assessments, which 83% deemed acceptable. Most respondents (87% of authors and 95% of editorial staff) thought RoB assessments were an improvement over past approaches to trial quality assessment. Most authors liked the standardized approach (81%) and the ability to provide quotes to support judgements (74%). A third of participants disliked the increased workload and found the wording describing RoB judgements confusing. The RoB domains reported to be the most difficult to assess were incomplete outcome data and selective reporting of outcomes. Authors expressed the need for more guidance on how to incorporate RoB assessments into meta-analyses and review conclusions. Based on this evaluation, recommendations were made for improvements to the RoB tool and the associated guidance. The implementation of these recommendations is currently underway.

CONCLUSIONS

Overall, respondents identified positive experiences and perceptions of the RoB tool. Revisions of the tool and associated guidance made in response to this evaluation, and improved provision of training, may improve implementation.

Effect modification by population dietary folate on the association between MTHFR genotype, homocysteine, and stroke risk: a meta-analysis of genetic studies and randomised trials

BACKGROUND

The MTHFR 677C→T polymorphism has been associated with raised homocysteine concentration and increased risk of stroke. A previous overview showed that the effects were greatest in regions with low dietary folate consumption, but differentiation between the effect of folate and small-study bias was difficult. A meta-analysis of randomised trials of homocysteine-lowering interventions showed no reduction in coronary heart disease events or stroke, but the trials were generally set in populations with high folate consumption. We aimed to reduce the effect of small-study bias and investigate whether folate status modifies the association between MTHFR 677C→T and stroke in a genetic analysis and meta-analysis of randomised controlled trials.

METHODS

We established a collaboration of genetic studies consisting of 237 datasets including 59,995 individuals with data for homocysteine and 20,885 stroke events. We compared the genetic findings with a meta-analysis of 13 randomised trials of homocysteine-lowering treatments and stroke risk (45,549 individuals, 2314 stroke events, 269 transient ischaemic attacks).

FINDINGS

The effect of the MTHFR 677C→T variant on homocysteine concentration was larger in low folate regions (Asia; difference between individuals with TT versus CC genotype, 3·12 μmol/L, 95% CI 2·23 to 4·01) than in areas with folate fortification (America, Australia, and New Zealand, high; 0·13 μmol/L, -0·85 to 1·11). The odds ratio (OR) for stroke was also higher in Asia (1·68, 95% CI 1·44 to 1·97) than in America, Australia, and New Zealand, high (1·03, 0·84 to 1·25). Most randomised trials took place in regions with high or increasing population folate concentrations. The summary relative risk (RR) of stroke in trials of homocysteine-lowering interventions (0·94, 95% CI 0·85 to 1·04) was similar to that predicted for the same extent of homocysteine reduction in large genetic studies in populations with similar folate status (predicted RR 1·00, 95% CI 0·90 to 1·11). Although the predicted effect of homocysteine reduction from large genetic studies in low folate regions (Asia) was larger (RR 0·78, 95% CI 0·68 to 0·90), no trial has evaluated the effect of lowering of homocysteine on stroke risk exclusively in a low folate region.

INTERPRETATION

In regions with increasing levels or established policies of population folate supplementation, evidence from genetic studies and randomised trials is concordant in suggesting an absence of benefit from lowering of homocysteine for prevention of stroke. Further large-scale genetic studies of the association between MTHFR 677C→T and stroke in low folate settings are needed to distinguish effect modification by folate from small-study bias. If future randomised trials of homocysteine-lowering interventions for stroke prevention are undertaken, they should take place in regions with low folate consumption.

FUNDING

Full funding sources listed at end of paper (see Acknowledgments).

Prognostic Importance of Anaemia in HIV Type-1-Infected Patients Starting Antiretroviral Therapy: Collaborative Analysis of Prospective Cohort Studies

Background

In HIV type-1-infected patients starting highly active antiretroviral therapy (HAART), the prognostic value of haemoglobin when starting HAART, and of changes in haemoglobin levels, are not well defined.

Methods

We combined data from 10 prospective studies of 12,100 previously untreated individuals (25% women). A total of 4,222 patients (35%) were anaemic: 131 patients (1.1%) had severe (<8.0 g/dl), 1,120 (9%) had moderate (male 8.0–<11.0 g/dl and female 8.0–<10.0 g/ dl) and 2,971 (25%) had mild (male 11.0–<13.0 g/dl and female 10.0–<12.0 g/dl) anaemia. We separately analysed progression to AIDS or death from baseline and from 6 months using Weibull models, adjusting for CD4+ T-cell count, age, sex and other variables.

Results

During 48,420 person-years of follow-up 1,448 patients developed at least one AIDS event and 857 patients died. Anaemia at baseline was independently associated with higher mortality: the adjusted hazard ratio (95% confidence interval) for mild anaemia was 1.42 (1.17–1.73), for moderate anaemia 2.56 (2.07–3.18) and for severe anaemia 5.26 (3.55–7.81). Corresponding figures for progression to AIDS were 1.60 (1.37–1.86), 2.00 (1.66–2.40) and 2.24 (1.46–3.42). At 6 months the prevalence of anaemia declined to 26%. Baseline anaemia continued to predict mortality (and to a lesser extent progression to AIDS) in patients with normal haemoglobin or mild anaemia at 6 months.

Conclusions

Anaemia at the start of HAART is an important factor for short- and long-term prognosis, including in patients whose haemoglobin levels improved or normalized during the first 6 months of HAART.

Graphical presentation of diagnostic information

BACKGROUND

Graphical displays of results allow researchers to summarise and communicate the key findings of their study. Diagnostic information should be presented in an easily interpretable way, which conveys both test characteristics (diagnostic accuracy) and the potential for use in clinical practice (predictive value).

METHODS

We discuss the types of graphical display commonly encountered in primary diagnostic accuracy studies and systematic reviews of such studies, and systematically review the use of graphical displays in recent diagnostic primary studies and systematic reviews.

RESULTS

We identified 57 primary studies and 49 systematic reviews. Fifty-six percent of primary studies and 53% of systematic reviews used graphical displays to present results. Dot-plot or box-and- whisker plots were the most commonly used graph in primary studies and were included in 22 (39%) studies. ROC plots were the most common type of plot included in systematic reviews and were included in 22 (45%) reviews. One primary study and five systematic reviews included a probability-modifying plot.

CONCLUSION

Graphical displays are currently underused in primary diagnostic accuracy studies and systematic reviews of such studies. Diagnostic accuracy studies need to include multiple types of graphic in order to provide both a detailed overview of the results (diagnostic accuracy) and to communicate information that can be used to inform clinical practice (predictive value). Work is required to improve graphical displays, to better communicate the utility of a test in clinical practice and the implications of test results for individual patients.

Prognostic importance of anaemia in HIV type-1-infected patients starting antiretroviral therapy: collaborative analysis of prospective cohort studies

BACKGROUND

In HIV type-1-infected patients starting highly active antiretroviral therapy (HAART), the prognostic value of haemoglobin when starting HAART, and of changes in haemoglobin levels, are not well defined.

METHODS

We combined data from 10 prospective studies of 12,100 previously untreated individuals (25% women). A total of 4,222 patients (35%) were anaemic: 131 patients (1.1%) had severe (<8.0 g/dl), 1,120 (9%) had moderate (male 8.0-<11.0 g/dl and female 8.0- < 10.0 g/dl) and 2,971 (25%) had mild (male 11.0- < 13.0 g/ dl and female 10.0- < 12.0 g/dl) anaemia. We separately analysed progression to AIDS or death from baseline and from 6 months using Weibull models, adjusting for CD4+ T-cell count, age, sex and other variables.

RESULTS

During 48,420 person-years of follow-up 1,448 patients developed at least one AIDS event and 857 patients died. Anaemia at baseline was independently associated with higher mortality: the adjusted hazard ratio (95% confidence interval) for mild anaemia was 1.42 (1.17-1.73), for moderate anaemia 2.56 (2.07-3.18) and for severe anaemia 5.26 (3.55-7.81). Corresponding figures for progression to AIDS were 1.60 (1.37-1.86), 2.00 (1.66-2.40) and 2.24 (1.46-3.42). At 6 months the prevalence of anaemia declined to 26%. Baseline anaemia continued to predict mortality (and to a lesser extent progression to AIDS) in patients with normal haemoglobin or mild anaemia at 6 months.

CONCLUSIONS

Anaemia at the start of HAART is an important factor for short- and long-term prognosis, including in patients whose haemoglobin levels improved or normalized during the first 6 months of HAART.

Prophylaxis of postoperative vomiting in children undergoing tonsillectomy: a systematic review and meta-analysis

Postoperative vomiting (POV) remains one of the commonest causes of significant morbidity after tonsillectomy in children. A variety of prophylactic anti-emetic interventions have been reported, but there has only been a limited systematic review in this patient group. A systematic search was performed by using Cochrane Controlled Trials Register, MEDLINE and EMBASE to identify double-blind, randomized, placebo-controlled trials of prophylactic anti-emetic interventions in children undergoing tonsillectomy, with or without adenoidectomy. The outcome of interest was POV in the first 24 h. Summary estimates of the effect of each prophylactic anti-emetic strategy were derived using fixed effect meta-analysis. Where appropriate, dose-response effects were estimated using logistic regression and 22 articles were identified. Good evidence was found for the prophylactic anti-emetic effect of dexamethasone [odds ratio (OR) 0.23, 95% CI 0.16-0.33], and the serotinergic antagonists ondansetron (OR 0.36, 95% CI 0.29-0.46), granisetron (OR 0.11, 95% CI 0.06-0.19), tropisetron (OR 0.15, 95% CI 0.06-0.35) and dolasetron (OR 0.25, 95% CI 0.1-0.59). Metoclopramide was also found to be efficacious (OR 0.51, 95% CI 0.34-0.77). There is not sufficient evidence to suggest that dimenhydrinate, perphenazine or droperidol, in the doses studied, are efficacious, nor were gastric aspiration or acupuncture. In conclusion, dexamethasone and the anti-serotinergic agents appear to be the most effective agents for the prophylaxis for POV in children undergoing tonsillectomy.

Vitamin A supplementation for reducing the risk of mother-to-child transmission of HIV infection

BACKGROUND

Mother-to-child transmission (MTCT) of HIV is the dominant mode of acquisition of HIV infection for children, currently resulting in about 1800 new paediatric HIV infections each day world-wide. This is one of several reviews assessing the available evidence for preventing HIV transmission from an HIV-infected woman to her child. The other reviews assess the effects of antiretroviral therapy, Caesarean section delivery, breast feeding, and vaginal lavage.

OBJECTIVES

To assess the effects of antenatal and intrapartum vitamin A supplementation, compared to an appropriate control group, on the risk of MTCT of HIV infection and infant and maternal mortality and morbidity, and the tolerability of vitamin A supplementation.

SEARCH STRATEGY

We searched the Cochrane Controlled Trials Register, Cochrane Pregnancy and Childbirth Register, PubMed, EMBASE, AIDSLINE, LILACS, AIDSTRIALS, and AIDSDRUGS, using standardised methodological filters for identifying trials. We also searched reference lists of identified articles, relevant editorials, expert opinions and letters to journal editors, and abstracts or proceedings of relevant conferences; and contacted subject experts, agencies, organisations, academic centres, and pharmaceutical companies. There were no language restrictions.

SELECTION CRITERIA

Randomised trials comparing vitamin A supplementation with no vitamin A supplementation in known HIV infected pregnant women. Trials had to include an estimate of the effect of vitamin A supplementation on MTCT of HIV and/or any other pre-specified adverse pregnancy outcome to be included.

DATA COLLECTION AND ANALYSIS

Two reviewers independently assessed trial eligibility and quality and extracted data. Odds ratios (OR) and 95% confidence intervals (CI) were estimated for binary data and pooled using a fixed effect (Mantel-Haenszel) method. Heterogeneity between studies was examined by graphical inspection of results followed by a chi-square test of homogeneity.

MAIN RESULTS

We identified five eligible trials, only two of which included an estimated of the effect of vitamin A supplementation on at least one of the pre-specified outcomes. Based on the two trials, with a total of 1813 participants, there is no evidence that vitamin A supplementation has an effect on MTCT of HIV (OR 1.09, 95% confidence interval (CI) 0.81 to 1.45). There is no evidence of heterogeneity between the trials (p = 0.37), and no evidence of an effect of vitamin A supplementation in HIV-infected pregnant women on stillbirths (OR 1.07, 95% CI 0.63 to 1.80), very preterm births, i.e. born less than 34 weeks gestation (OR 0.86, 95% CI 0.57 to 1.31), all preterm births, i.e. born less than 37 weeks gestation (OR 0.88, 95% CI 0.68 to 1.13), low birth weight, i.e. weighing less than 2500g (OR 0.86, 95% CI 0.64 to 1.17), very low birthweight, i.e. weighing less than 2000g (OR 0.71, 95% CI 0.40 to 1.28), and postpartum CD4 levels (weighted mean difference -4.00, 95% CI -51.06 to 43.06). The effect of vitamin A on maternal mortality could not be assesssed, as there were only three maternal deaths.

IMPLICATIONS FOR PRACTICE

At the present time there is no conclusive evidence that the antenatal and intrapartum use of vitamin A supplementation to reduce MTCT of HIV and adverse pregnancy outcomes among HIV-infected pregnant women should be recommended.

IMPLICATIONS FOR RESEARCH

The current review will be updated as soon as data from ongoing studies become available. This review and the review in progress on vitamin A supplementation in pregnant women of seronegative/unknown HIV status (Kulier 2002) should be considered together in order to shed more light on the effect of vitamin A supplementation on non-HIV related adverse pregnancy outcomes.

Vaginal disinfection during labour for reducing the risk of mother-to-child transmission of HIV infection

BACKGROUND

Mother-to-child transmission (MTCT) of HIV infection is one of the most tragic consequences of the HIV epidemic, especially in resource-limited countries, resulting in about 650 000 new paediatric HIV infections each year world-wide. The paediatric HIV epidemic threatens to seriously undermine decade-old child survival programmes. This is one of several reviews assessing the available evidence for preventing HIV transmission from an HIV-infected woman to her child. The other reviews assess the effects of antiretroviral therapy, Caesarean section delivery, breast feeding, and vitamin A supplementation.

OBJECTIVES

To estimate the effect of vaginal lavage on the risk of MTCT of HIV and infant and maternal mortality and morbidity, as well as tolerability of vaginal lavage in HIV infected women.

SEARCH STRATEGY

We searched the Cochrane Controlled Trials Register, Cochrane Pregnancy and Childbirth Register, PubMed, EMBASE, AIDSLINE, LILACS, AIDSTRIALS, and AIDSDRUGS, using standardised methodological filters for identifying trials. We also searched reference lists of identified articles, relevant editorials, expert opinions and letters to journal editors, and abstracts and proceedings of relevant conferences, and contacted subject experts and pharmaceutical companies. There were no language restrictions.

SELECTION CRITERIA

Randomised trials or clinical trials comparing vaginal disinfection during labour with placebo or no treatment, in known HIV infected pregnant women. Trials had to include an estimate of the effect of vaginal disinfection on MTCT of HIV and/or any other pre-specified adverse pregnancy outcome to be included.

DATA COLLECTION AND ANALYSIS

Two reviewers independently assessed trial eligibility and quality, and extracted data.

MAIN RESULTS

Only one low quality trial included an estimate of the effect of vaginal disinfection on at least one pre-specified outcome. There was no evidence of an effect of vaginal disinfection on MTCT of HIV (odds ratio 0.93, 95% confidence interval 0.63 to 1.38), and no information was available on the other pre-specified outcomes.

REVIEWER'S CONCLUSIONS

There is a need for well-designed randomised controlled trials to estimate the effects of vaginal disinfection on MTCT of HIV.

Estimation of the incidence of stroke using a capture-recapture model including covariates

BACKGROUND

Capture-recapture is often used to assess completeness of a register. However, the usual two-source model relies on assumptions of independence of sources and equality of capture probability which are rarely satisfied in epidemiology. An alternative is to include covariates in capture-recapture models.

METHODS

We use capture-recapture models including covariates to estimate incidence of stroke in South London. We estimate ascertainment-adjusted age-standardized incidence rates, and calculate confidence intervals for incidence which allow for the uncertainty in estimation of the total number of cases.

RESULTS

The crude capture-recapture model (including no covariates) underestimated the number of non-fatal strokes. Demographic and stroke severity variables were associated with the probability of capture. Including covariates led to more plausible results for fatal and non-fatal strokes, and suggested that the stroke register was 88% complete. Adjusting for under-ascertainment increased the estimated incidence from 1.31 (95% CI : 1.21-1.42) to 1.49 (95% CI : 0.38-2.60) per 1000 people.

CONCLUSIONS

Incidence and age-standardized incidence can be calculated using data from an incomplete register. However, sparse strata can lead to wide confidence intervals for adjusted rates. Cost-effectiveness of routine registers might be increased by using the combination of sources and covariates which most accurately estimates the total number of cases, rather than by aiming for 100% completeness.

A new method for predicting recovery after stroke

BACKGROUND AND PURPOSE

Several prognostic factors have been identified for outcome after stroke. However, there is a need for empirically derived models that can predict outcome and assist in medical management during rehabilitation. To be useful, these models should take into account early changes in recovery and individual patient characteristics. We present such a model and demonstrate its clinical utility.

METHODS

Data on functional recovery (Barthel Index) at 0, 2, 4, 6, and 12 months after stroke were collected prospectively for 299 stroke patients at 2 London hospitals. Multilevel models were used to model recovery trajectories, allowing for day-to-day and between-patient variation. The predictive performance of the model was validated with an independent cohort of 710 stroke patients.

RESULTS

Urinary incontinence, sex, prestroke disability, and dysarthria affected the level of outcome after stroke; age, dysphasia, and limb deficit also affected the rate of recovery. Applying this to the validation cohort, the average difference between predicted and observed Barthel Index was -0.4, with 90% limits of agreement from -7 to 6. Predicted Barthel Index lay within 3 points of the observed Barthel Index on 49% of occasions and improved to 69% when patients' recovery histories were taken into account.

CONCLUSIONS

The model predicts recovery at various stages of rehabilitation in ways that could improve clinical decision making. Predictions can be altered in light of observed recovery. This model is a potentially useful tool for comparing individual patients with average recovery trajectories. Patients at elevated risk could be identified and interventions initiated.

Dietary antioxidants and asthma in adults: population-based case-control study

A protective role for dietary antioxidants in asthma has been proposed. However, epidemiological evidence to implicate antioxidant vitamins is weak, and data on the role of flavonoid-rich foods and antioxidant trace elements are lacking. We carried out a population-based case-control study in South London, UK, to investigate whether asthma is less common and less severe in adults who consume more dietary antioxidants. Participants were aged 16-50 yr and registered with 40 general practices. Asthma was defined by positive responses to a standard screening questionnaire in 1996, and complete information about usual diet was obtained by food frequency questionnaire from 607 cases and 864 controls in 1997. After controlling for potential confounding factors and total energy intake, apple consumption was negatively associated with asthma (odds ratio [OR] per increase in frequency group 0.89 [95% confidence interval [CI]: 0.82 to 0.97]; p = 0.006). Intake of selenium was also negatively associated with asthma (OR per quintile increase 0.84 [0.75 to 0.94]; p = 0.002). Red wine intake was negatively associated with asthma severity. The associations between apple and red wine consumption and asthma may indicate a protective effect of flavonoids. The findings for dietary selenium could have implications for health policy in Britain where intake has been declining.

Uses and abuses of meta-analysis

Meta-analysis, the statistical combination of results from several studies to produce a single estimate of the effect of a treatment, continues to attract controversy. We illustrate the potentials and pitfalls of meta-analysis of controlled clinical trials. Cumulative meta-analysis demonstrates that this technique could prevent delays in the introduction of effective treatments. Meta-analyses are, however, liable to numerous biases both at the level of the individual trial ('garbage in, garbage out') and the dissemination of trial results (publication bias). We argue that meta-analysis should be performed only within the framework of systematic reviews--that is, reviews prepared using a systematic approach to minimise bias and address the combinability of studies.

Funnel plots for detecting bias in meta-analysis: guidelines on choice of axis

Asymmetry in funnel plots may indicate publication bias in meta-analysis, but the shape of the plot in the absence of bias depends on the choice of axes. We evaluated standard error, precision (inverse of standard error), variance, inverse of variance, sample size and log sample size (vertical axis) and log odds ratio, log risk ratio and risk difference (horizontal axis). Standard error is likely to be the best choice for the vertical axis: the expected shape in the absence of bias corresponds to a symmetrical funnel, straight lines to indicate 95% confidence intervals can be included and the plot emphasises smaller studies which are more prone to bias. Precision or inverse of variance is useful when comparing meta-analyses of small trials with subsequent large trials. The use of sample size or log sample size is problematic because the expected shape of the plot in the absence of bias is unpredictable. We found similar evidence for asymmetry and between trial variation in a sample of 78 published meta-analyses whether odds ratios or risk ratios were used on the horizontal axis. Different conclusions were reached for risk differences and this was related to increased between-trial variation. We conclude that funnel plots of meta-analyses should generally use standard error as the measure of study size and ratio measures of treatment effect.

Multilevel growth curve models with covariate effects: application to recovery after stroke

In measuring the progression of, or recovery from, a disease an individual's outcome may be assessed on a number of occasions. A model of the relationship between outcome and time since disease occurred which accounts for patient characteristics could be used to describe patterns of recovery, to predict outcome for a patient, or to evaluate health interventions. We use multilevel models to analyse such data, focusing on the choice of powers of time both for mean outcome and covariate effects. We give equations for predicted outcome and corresponding standard errors (i) based only on baseline characteristics, and (ii) by conditioning on previous outcomes for an individual. In a study of 331 stroke patients, outcome was measured approximately 0, 2,4,6 and 12 months after stroke. Patient characteristics included age, sex, and pre-stroke handicap, together with stroke-severity indicators (presence of limb deficit, dysphasia, dysarthria or incontinence). Of these, only the effects of age, dysphasia and presence of deficit varied with time. Conditioning on previous observations improved the accuracy of predictions. The outcome variable clearly had a skewed distribution, and the model residuals showed evidence of non-Normality. We discuss alternative models for non-Normal data, and show that, here, the standard (Normal errors) multilevel model gives equivalent parameter estimates and predictions to those obtained from alternative models.

Inequalities in rates of gonorrhoea and chlamydia between black ethnic groups in south east London: cross sectional study

OBJECTIVES

To examine differences in population based rates of gonorrhoea and chlamydia between black ethnic groups in Lambeth, Southwark and Lewisham Health Authority.

METHODS

Episodes of gonorrhoea or chlamydia recorded among attenders at 11 genitourinary clinics in south and central London from 1 January 1994 to 31 December 1995 were retrieved. Complete data on chlamydia were only available for women. Ethnic group was assigned according to census categories--white, black Caribbean, black African, black other, Asian, or other. We calculated yearly incidence rates for episodes of gonorrhoea and chlamydia in residents of Lambeth, Southwark and Lewisham Health Authority. Random effects Poisson regression models were used to examine associations between infection rates and age, ethnic group, and material deprivation.

RESULTS

During the study period there were 1996 episodes of gonorrhoea in men and women and 1376 episodes of chlamydia in women with complete data. For both infections rates among individuals from black Caribbean and black other ethnic groups were markedly higher than among black Africans. In men, the gonorrhoea rate among black Caribbean 20-24 year olds was 2348 (95% CI 1965 to 2831) episodes per 100,000 compared with 931 (95% CI 690 to 1288) in black African men and 111 (95% CI 100 to 124) per 100,000 in white men of the same age. Among women gonorrhoea rates were highest in black Caribbean 15-19 year olds (2612, 95% CI 2161 to 3190 per 100,000). In contrast, rates in black African women of the same age (331, 95% CI 154 to 846 per 100,000) were similar to those of white women (222, 95% CI 163 to 312). Chlamydia rates were also highest in black Caribbean 15-19 year old women (4579, 95% CI 3966 to 5314 per 100,000), compared with 1286 (95% CI 907 to 1888) in black African and 433 (95% CI 349 to 544) per 100,000 white women. Controlling for material deprivation and age only attenuated differences in rates between ethnic groups slightly.

CONCLUSIONS

There are marked differences in rates of gonorrhoea and chlamydia between different black ethnic groups, with higher rates in black Caribbeans than black Africans. This study supports the hypothesis that assortative sexual mixing patterns can restrict epidemics of sexually transmitted infections within ethnic groups. Differences in disease occurrence between black ethnic groups should be explored before combining data, even when numbers of episodes are small.

Publication and related bias in meta-analysis: power of statistical tests and prevalence in the literature

Publication and selection biases in meta-analysis are more likely to affect small studies, which also tend to be of lower methodological quality. This may lead to "small-study effects," where the smaller studies in a meta-analysis show larger treatment effects. Small-study effects may also arise because of between-trial heterogeneity. Statistical tests for small-study effects have been proposed, but their validity has been questioned. A set of typical meta-analyses containing 5, 10, 20, and 30 trials was defined based on the characteristics of 78 published meta-analyses identified in a hand search of eight journals from 1993 to 1997. Simulations were performed to assess the power of a weighted regression method and a rank correlation test in the presence of no bias, moderate bias or severe bias. We based evidence of small-study effects on P < 0.1. The power to detect bias increased with increasing numbers of trials. The rank correlation test was less powerful than the regression method. For example, assuming a control group event rate of 20% and no treatment effect, moderate bias was detected with the regression test in 13.7%, 23.5%, 40.1% and 51.6% of meta-analyses with 5, 10, 20 and 30 trials. The corresponding figures for the correlation test were 8.5%, 14.7%, 20.4% and 26.0%, respectively. Severe bias was detected with the regression method in 23.5%, 56.1%, 88.3% and 95.9% of meta-analyses with 5, 10, 20 and 30 trials, as compared to 11.9%, 31.1%, 45.3% and 65.4% with the correlation test. Similar results were obtained in simulations incorporating moderate treatment effects. However the regression method gave false-positive rates which were too high in some situations (large treatment effects, or few events per trial, or all trials of similar sizes). Using the regression method, evidence of small-study effects was present in 21 (26.9%) of the 78 published meta-analyses. Tests for small-study effects should routinely be performed in meta-analysis. Their power is however limited, particularly for moderate amounts of bias or meta-analyses based on a small number of small studies. When evidence of small-study effects is found, careful consideration should be given to possible explanations for these in the reporting of the meta-analysis.

Mechanisms of risk in preterm low-birthweight infants

The periodontal diseases share many common risk factors with preterm low birth weight. Examples are, age, socioeconomic status and smoking (Fig. 5). Studies to date have only shown an association between the two conditions, and this does not indicate a causal relationship. However, since the inflammatory mediators that occur in the periodontal diseases, also play an important part in the initiation of labor, there are plausible biological mechanisms that could link the two conditions. The challenge for the future is to characterize the nature of the factors that predispose a mother to give birth prematurely to infants less than 2500 g and to assign relative probabilities to each. Studies are taking place in many parts of the world to determine the probability of a preterm low-birth-weight outcome, the interdependence of the factors that contribute to a birth event and possible casual relationships between these factors. Further information about the details of the effects of maternal infection will come from intervention studies, animal studies and more detailed examination of the mechanisms.

Frequent paracetamol use and asthma in adults

BACKGROUND

The pulmonary antioxidant glutathione may limit airway inflammation in asthma. Since paracetamol (acetaminophen) depletes the lung of glutathione in animals, a study was undertaken to investigate whether frequent use in humans was associated with asthma.

METHODS

Information was collected on the use of analgesics as part of a population based case-control study of dietary antioxidants and asthma in adults aged 16-49 years registered with 40 general practices in Greenwich, South London. The frequency of use of paracetamol and aspirin was compared in 664 individuals with asthma and in 910 without asthma. Asthma was defined by positive responses to questions about asthma attacks, asthma medication, or waking at night with shortness of breath. The association between analgesic use and severity of disease amongst asthma cases, as measured by a quality of life score, was also examined.

RESULTS

Paracetamol use was positively associated with asthma. After controlling for potential confounding factors the odds ratio for asthma, compared with never users, was 1.06 (95% CI 0.77 to 1.45) in infrequent users (<monthly), 1.22 (0.87 to 1.72) in monthly users, 1. 79 (1.21 to 2.65) in weekly users, and 2.38 (1.22 to 4.64) in daily users (p (trend) = 0.0002). This association was present in users and non-users of aspirin and was stronger when cases with more severe disease were compared with controls; amongst cases increasing paracetamol use was associated with more severe disease. Frequency of aspirin use was not associated with asthma when cases as a whole were compared with controls, nor with severity of asthma amongst cases. Frequent paracetamol use was positively associated with rhinitis, but aspirin use was not.

CONCLUSIONS

Frequent use of paracetamol may contribute to asthma morbidity and rhinitis in adults.

Influence of surgeon's experience on speech outcome in cleft lip and palate

Relation between the number of cases treated and outcome, particularly in low-incidence anomalies, are difficult to identify and rectify. We show clear advantages in speech outcomes for children born with a cleft lip and palate who are operated on by surgeons who do large numbers of these procedures.

Clustered randomised trial of an intervention to improve the management of asthma: Greenwich asthma study

OBJECTIVES

To evaluate the effectiveness of an asthma resource centre in improving treatment and quality of life for asthmatic patients.

DESIGN

Community based randomised controlled trial.

SETTING

41 general practices in Greenwich with a practice nurse.

SUBJECTS

All registered patients aged 15-50 years.

INTERVENTION

Nurse specialists in asthma who educated and supported practice nurses, who in turn educated patients in the management of asthma according to the British Thoracic Society's guidelines.

MAIN OUTCOME MEASURES

Quality of life of asthmatic patients, attendance at accident and emergency departments, admissions to local hospitals, and steroid prescribing by general practitioners.

RESULTS

Of 24 400 patients randomly selected and surveyed in 1993, 12 238 replied; 1621 were asthmatic of whom 1291 were sent a repeat questionnaire in 1996 and 780 replied. Of 24 400 patients newly surveyed in 1996, 10 783 (1616 asthmatic) replied. No evidence was found for an improvement in asthma related quality of life among newly surveyed patients in intervention practices compared with control practices. Neither was there evidence of an improvement in other measures of the quality of asthma care. Weak evidence was found for an improvement in quality of life in intervention practices among asthmatics registered with study practices in 1993 and followed up in 1996. Neither attendances at accident and emergency departments nor admissions for asthma showed any tendency to diverge in intervention and control practices over the study period. Steroid prescribing rates rose steadily during the study period. The average annual increase in steroid prescribing was 3% per year higher in intervention than control practices (95% confidence interval -1% to 6%, P=0.10).

CONCLUSIONS

This model of service delivery is not effective in improving the outcome of asthma in the community. Further development is required if cost effective management of asthma is to be introduced.

Birth weight, body mass index and asthma in young adults

BACKGROUND

Impaired fetal growth may be a risk factor for asthma although evidence in children is conflicting and there are few data in adults. Little is known about risk factors which may influence asthma in late childhood or early adult life. Whilst there are clues that fatness may be important, this has been little studied in young adults. The relations between birth weight and childhood and adult anthropometry and asthma, wheeze, hayfever, and eczema were investigated in a nationally representative sample of young British adults.

METHODS

A total of 8960 individuals from the 1970 British Cohort Study (BCS70) were studied. They had recently responded to a questionnaire at 26 years of age in which they were asked whether they had suffered from asthma, wheeze, hayfever, and eczema in the previous 12 months. Adult body mass index (BMI) was calculated from reported height and weight.

RESULTS

The prevalence of asthma at 26 years fell with increasing birth weight. After controlling for potential confounding factors, the odds ratio comparing the lowest birth weight group (<2 kg) with the modal group (3-3.5 kg) was 1.99 (95% CI 0.96 to 4.12). The prevalence of asthma increased with increasing adult BMI. After controlling for birth weight and other confounders, the odds ratio comparing highest with lowest quintile was 1.72 (95% CI 1.29 to 2.29). The association between fatness and asthma was stronger in women; odds ratios comparing overweight women (BMI 25-29.99) and obese women (BMI >/=30) with those of normal weight (BMI <25) were 1.51 (95% CI 1.11 to 2.06) and 1.84 (95% CI 1. 19 to 2.84), respectively. The BMI at 10 years was not related to adult asthma. Similar associations with birth weight and adult BMI were present for wheeze but not for hayfever or eczema.

CONCLUSIONS

Impaired fetal growth and adult fatness are risk factors for adult asthma.

Capture-recapture models including covariate effects

Capture-recapture methods are used to estimate the incidence of a disease, using a multiple-source registry. Usually, log-linear methods are used to estimate population size, assuming that not all sources of notification are dependent. Where there are categorical covariates, a stratified analysis can be performed. The multinomial logit model has occasionally been used. In this paper, the authors compare log-linear and logit models with and without covariates, and use simulated data to compare estimates from different models. The crude estimate of population size is biased when the sources are not independent. Analyses adjusting for covariates produce less biased estimates. In the absence of covariates, or where all covariates are categorical, the log-linear model and the logit model are equivalent. The log-linear model cannot include continuous variables. To minimize potential bias in estimating incidence, covariates should be included in the design and analysis of multiple-source disease registries.

Delayed-type hypersensitivity to Mycobacterium leprae soluble antigens as a test for infection with the leprosy bacillus

BACKGROUND

Mycobacterium leprae (M. leprae) soluble antigen (MLSA) reagents have been developed with the aim of finding a reagent, comparable to tuberculin, which could identify individuals infected with the leprosy bacillus. They have yet to be evaluated fully in human populations.

METHODS

More than 15000 individuals living in a leprosy endemic area of northern Malawi were skin tested with one of five batches of MLSA prepared using two different protocols. The main difference in preparation was the introduction of a high G centrifugation step in the preparation of the last three ('second-generation') batches.

RESULTS

The prevalence of skin-test positivity (delayed-type hypersensitivity (DTH)) and association with the presence of a BCG scar were greater for first (batches A6, A22) than second (batches AB53, CD5, CD19) generation reagents. The association of positivity with M. leprae infection was investigated by comparing results among known (household) contacts of leprosy cases, and among newly diagnosed leprosy patients with those in the general population. While positivity to 'first-generation' antigens appeared to be associated with M. leprae infection, positivity to later antigens was unrelated either to exposure to leprosy cases or presence of leprosy disease. There were geographical differences in the prevalence of DTH to the various batches, probably reflecting exposure to various mycobacteria in the environment.

CONCLUSIONS

Our results suggest that the 'second-generation' batches have lost antigens that can detect M. leprae infections, but that they retain one or more antigens which are shared between M. leprae and environmental mycobacteria. Natural exposure to these both sensitizes individuals and provides natural protection against M. leprae infection or disease. Identification of antigens present in these groups of skin test reagents may assist in production of improved skin test reagents.

The East London Study of Maternal Chronic Periodontal Disease and Preterm Low Birth Weight Infants: study design and prevalence data

The influence of subject-based and environmental factors on the balance between the subgingival microbial challenge and the host response in periodontal diseases illustrates the intimate link between oral and systemic health. From this stems the hypothesis that the persistent Gram-negative challenge and associated inflammatory sequelae in periodontal disease may have consequences extending beyond the periodontal tissues themselves. This paper addresses the design of a case-control study to examine the relationship between preterm low birth weight (PLBW) and maternal periodontal disease. We present preliminary data on the prevalence of these 2 conditions in a group of mothers at the Royal Hospitals Trust, London, U.K. Cases are defined as mothers delivering an infant weighing less than 2,500g before 37 weeks gestation and controls as mothers delivering an infant of more than 2,500g after 38 weeks. We estimated that a study involving 800 mothers (1:3 case:control) should have sufficient power to detect an association with a minimum odds ration of 3 at the 5% significance level. Demographic details of 177 subjects demonstrated that they were representative of the local population, and the prevalence of PLBW was within the expected range. However, the extent and severity of periodontal disease were higher than predicted and may have reflected elevations in gingival inflammation associated with pregnancy. The final outcome of the study should help determine the need for further interventionist studies to demonstrate a causal relationship between periodontal disease and PLBW, as well as provide information on the prevalence of periodontal diseases in this study population.

Birth weight, childhood lower respiratory tract infection, and adult lung function

BACKGROUND

Historical cohort studies in England have found that impaired fetal growth and lower respiratory tract infections in early childhood are associated with lower levels of lung function in late adult life. These relations are investigated in a similar study in Scotland.

METHODS

In 1985-86 a follow up study was carried out of 1070 children who had been born in St Andrew's from 1921 to 1935 and followed from birth to 14 years of age by the Mackenzie Institute for Medical Research. Recorded information included birth weight and respiratory illnesses. The lung function of 239 of these individuals was measured.

RESULTS

There was no association between birth weight and lung function. Pneumonia before two years of age was associated with a difference in mean forced expiratory volume in one second (FEV1) of -0.39 litres (95% confidence interval (CI) -0.67, -0.11; p = 0.007) and in mean forced vital capacity (FVC) of -0.60 litres (95% CI -0.92, -0.28; p < 0.001), after controlling for age, sex, height, smoking, type of spirometer, and other illnesses before two years. Similar reductions were seen in men and women. Bronchitis before two years was associated with smaller deficits in FEV1 and FVC. Asthma or wheeze at two years and older and cough after five years were also associated with a reduction in FEV1.

CONCLUSIONS

The relation between impaired fetal growth and lower lung function in late adult life seen in previous studies was not confirmed in this cohort. The deficits in FEV1 and FVC associated with pneumonia and bronchitis in the first two years of life are consistent with a causal relation.

Does the efficacy of BCG decline with time since vaccination?

OBJECTIVE

To investigate whether the protective efficacy of bacille Calmette-Guérin (BCG) against tuberculosis decreases with time since vaccination.

DESIGN

A quantitative review of all 10 randomized trials of BCG against tuberculosis in purified protein derivative (PPD)-negative individuals, that presented data for discrete periods. For each trial, we derived log rate ratios for the annual change in the efficacy of BCG. We also compared efficacy in the first two years, and the first 10 years, to that in the rest of the trial.

RESULTS

There was considerable heterogeneity between trials in the annual change in the efficacy of BCG. In seven efficacy decreased overtime, while in three it increased. Average annual change in efficacy was not related to overall efficacy. Efficacy also varied between trials in the first two years after vaccination, at more than two years after vaccination and in the first ten years after vaccination. However the variation in efficacy between trials more than 10 years after vaccination was not statistically significant (P = 0.26). We therefore calculated that the average efficacy more than 10 years after vaccination was 14% (95% confidence interval -9% to 32%).

CONCLUSION

BCG protection can wane with time since vaccination. There is no good evidence that BCG provides protection more than 10 years after vaccination.

Household and dwelling contact as risk factors for leprosy in northern Malawi

Data on household and dwelling contact with known leprosy cases were available on more than 80,000 initially disease-free individuals followed up during the 1980s in a rural district of northern Malawi. A total of 331 new cases of leprosy were diagnosed among them. Individuals recorded as living in household or dwelling contact with multibacillary patients at the start of follow-up were at approximately five- to eightfold increased risk of leprosy, respectively, compared with individuals not living in such households or dwellings. Individuals living in household or dwelling contact with paucibacillary cases were both at approximately twofold increased risk. The higher risk associated with multibacillary contact and the fact that dwelling contact entailed a greater risk than household contact if the association was with multibacillary, but not with paucibacillary, disease suggest that paucibacillary cases may not themselves be sources of transmission, but rather just markers that a household has had contact with some (outside) source of infection. When household contact was considered alone, the risks of disease were appreciably higher for younger than for older contacts and for male compared with female contacts. Despite the elevated risk of leprosy associated with household or dwelling contact, only 15% of all incidence cases arose among recognized household contacts. Given the dynamic nature of household membership and consequent misclassification of contact status, the true contribution to overall incidence of contact within household or dwelling settings is likely to be much higher than this, perhaps 30% or higher. Considering the predilection of males for infectious multibacillary forms of the disease, the transmission of Mycobacterium leprae at an early age, in particular to males, may be of particular importance for the persistence of leprosy in endemic communities. Although residential contact with a multibacillary case is the strongest known determinant of leprosy risk, the vast majority of such contacts never manifest disease, which indicates a crucial role for genetic and/or environmental factors in the transmission of M. leprae infection and/or the pathogenesis of clinical leprosy.

Comparison of bacillary indexes in slit-skin smears, skin and nerve biopsies; a study from Malawi

Data analyzed in this paper were collected within the framework of the Lepra Evaluation Project, an epidemiological study of leprosy in Karonga District, northern Malawi. For 212 patients information on the number of skin lesions, slit-skin smear and skin biopsy results were available. Among 61 patients with a single lesion none were slit-skin-smear positive and two had bacilli detected in skin biopsies. In contrast, among 119 patients with four or more lesions 34 (28.6%) versus 59 (49.6%) had bacilli detectable in slit-skin smears or skin biopsies, respectively. In a further 47 patients skin biopsy results could be compared with split-nerve biopsy results. In 20 of 47 patients the bacterial indexes (BIs) were identical in skin and nerve biopsies, while in 26 of 47 patients the BIs were higher in nerve than in skin biopsies. This difference, which is consistent with several other studies in the literature, provides an insight into the pathogenesis of leprosy.

Use of an internal standard in subtraction radiography to assess initial periodontal bone changes

OBJECTIVE

To assess the use of an internal reference when performing histogram analyses in digital subtraction images and to determine the ability of the method to detect initial bone lesions.

METHODS

Fifty-one Royal Air Force recruits had standardized vertical bitewing radiographs and clinical assessment of attachment level recorded annually over three years. Subtraction analyses of crestal bone changes at the mesial surface of the upper right first molar were compared with changes at the mesial surface of the same tooth. Changes over the periods from age 17 to 18 years and age 18 to 20 years were monitored using two subtraction procedures.

RESULTS

Reproducibility studies revealed that Pearson correlations between duplicate measurements of the test site alone (16 crest; r = 0.74) were lower than those between duplicate measurements of changes where misalignment was controlled for, either as a difference (16 crest-16 tooth; r = 0.93) or ratio (16 crest/16 tooth; r = 0.93). We used the differences between the mean subtraction density for 16 crest and 16 tooth as our measure of change in bone density. For the duplicate measurements, the standard deviation of these differences was 3.9: a difference of +/- 7.8 was therefore taken as a threshold value for evidence of real change. Bone gain was noted between the ages of 17 and 18 years (16/21 subjects), but some early bone loss was seen between 18 and 20 years (12/21 subjects), with four subjects showing changes significantly greater than the method error. There were no associations between the clinical and radiographic observations.

CONCLUSION

Use of a control site in subtraction radiography improves the reproducibility; such systems can detect small changes in alveolar bone which may assist in early diagnosis of the initial periodontal lesion which may precede observable clinical changes.

Does bacille Calmette-Guérin scar size have implications for protection against tuberculosis or leprosy?

SETTING

Total population study in Karonga District, northern Malawi, in which the overall vaccine efficacy of bacille Calmette-Guérin (BCG) has been found to be -7% against tuberculosis and 54% against leprosy.

OBJECTIVE

To examine the relationship between BCG scar size and protection against tuberculosis and leprosy.

DESIGN

Cohort study in which 85,134 individuals were screened for tuberculosis and 82,265 for leprosy between 1979 and 1984, and followed up between 1986 and 1989.

RESULTS

Of the BCG scar positive individuals whose scars were measured, 31/3 2471 were later identified with tuberculosis and 81/31 879 with leprosy. In 19,114 individuals, of whom 17 developed tuberculosis, tuberculin induration was measured at first examination. Mean scar sizes increased with increasing tuberculin induration in all except the oldest individuals. Mean scar sizes were lowest in individuals aged < 10 years, highest in individuals aged 10-29 years and intermediate in older individuals. There was some evidence (P = 0.08) for an increase in tuberculosis risk with increasing scar size, which probably reflects the known correlation between scar size and tuberculin status at the time of vaccination. There was no clear association between BCG scar size and leprosy incidence.

CONCLUSIONS

We find no evidence that increased BCG scar size is a correlate of vaccine-induced protective immunity against either tuberculosis or leprosy.

Geographic determinants of leprosy in Karonga District, Northern Malawi

BACKGROUND

Geographical differences in leprosy risk are not understood, but may provide clues about the natural history of the disease. We report an analysis of the geographical distribution of leprosy in Karonga District, a rural area of Northern Malawi, between 1979 and 1989.

METHODS

Cohort study of the incidence of leprosy based on two total population surveys. Area of residence was determined using aerial photographs, which allowed identification of households, as well as location of roads, rivers and the lake shore.

RESULTS

Incidence rates were between two and three times higher in the north compared to the south of the district, and lowest in the semi-urban district capital. The most obvious environmental difference between these regions is the north's higher rainfall and more fertile soil. There was no overall association between leprosy incidence and population density, although highest rates were observed in the least densely populated areas. Looking at the entire district, incidence rates increased with increasing distance from a main road, but declined with increasing distance from a river or from the shore of Lake Malawi. The negative association with proximity to rivers may reflect the larger number of rivers in the north of the district. Apparent differences in incidence rates between groups speaking different languages reflected confounding by area of residence.

CONCLUSIONS

There is a marked variation, not explained by socioeconomic or cultural factors, in the incidence of leprosy within Karonga District. Our results are consistent with a theme in the literature associating the environment, particularly proximity to water, with leprosy.

Testing for antibody to human immunodeficiency virus type 1 in a population in which mycobacterial diseases are endemic

During a large epidemiologic study in the Karonga District of northern Malawi, serum samples from 139 patients with incident leprosy, 124 with newly diagnosed leprosy, 277 patients with incident tuberculosis, and 2296 controls were tested for antibodies to human immunodeficiency virus. Sera were tested according to a four-test protocol using two ELISAs and two particle agglutination assays. Overall, 188 samples were considered positive, 2634 were considered negative, and 14 were indeterminate. All 18 available positive specimens from leprosy patients, a random sample of 14 positive specimens from tuberculosis patients, and 15 positive specimens from controls were tested by Western blot. There was no evidence of substantial numbers of ELISA false-positives in any patient group or among controls.

Patterns of initial and acquired antituberculosis drug resistance in Karonga District, Malawi

There is concern that drug-resistant tuberculosis is increasing and may be concentrated among HIV-positive patients. Little information is available from developing countries, where surveillance studies are often unable to distinguish resistance in previously untreated patients (initial resistance) from resistance acquired following drug therapy, and where information on the HIV status of the patients is rare. Initial resistance patterns reflect the strains being transmitted in the community. We have studied patterns of resistance in northern Malawi, where the Lepra Evaluation Project has been collecting data on drug resistance since 1986. Initial drug sensitivity results were available for 373 new cases of tuberculosis. Initial resistance to at least one drug was found in 44 of these patients (11.8%, 95% CI 8.5-15.1): 13 were resistant to streptomycin alone, 13 to isoniazid alone, and 17 to more than one drug. Only 3 patients showed initial rifampicin resistance-1 in isolation, 1 in combination with streptomycin, and 1 with triple resistance. Drug resistance was not related to age, sex, or HIV status of the patient and there was no evidence of any increase over the period studied. There was no evidence of geographic clustering of the resistant strains, or of any increased risk of resistant strains in households with previous tuberculosis cases. Acquired resistance during follow-up was found in 5 of 329 patients with documented initially fully sensitive strains. 5 patients with initial resistance seemed to show reversion to sensitivity. The absence of an increase in drug resistance, despite an increase in tuberculosis cases over the period, is encouraging for the control programme. It emphasises the need to collect information from many areas before assuming that increases in antituberculosis drug resistance are occurring worldwide.

Epidemiological aspects of relapses in leprosy

Life table methods in which the cumulative probability of relapse in successive periods is calculated are preferable to the presentation of overall relapse rates. Their use facilitates the comparison of relapse rates and trends from different studies independent of duration of follow-up. Results from various studies including data from Malawi indicate that, (1) unlike after dapsone monotherapy, the cumulative probability of relapse in multibacillary patients is near to zero after WHO/MDT if strict definitions of relapse are used and, (2) the cumulative probability of relapse may approach 5% in paucibacillary patients 10 years after completion of WHO/MDT. On the whole, the epidemiological relevance of relapses is insignificant and future treatment regimens should be evaluated concerning their efficacy in preventing disabilities rather than relapses.