Professional assistance during birth and maternal mortality in two Indonesian districts.

OBJECTIVE: To examine determinants of maternal mortality and assess the effect of programmes aimed at increasing the number of births attended by health professionals in two districts in West Java, Indonesia. METHODS: We used informant networks to characterize all maternal deaths, and a capture-recapture method to estimate the total number of maternal deaths. Through a survey of recent births we counted all midwives practising in the two study districts. We used case-control analysis to examine determinants of maternal mortality, and cohort analysis to estimate overall maternal mortality ratios. FINDINGS: The overall maternal mortality ratio was 435 per 100,000 live births (95% confidence interval, CI: 376-498). Only 33% of women gave birth with assistance from a health professional, and among them, mortality was extremely high for those in the lowest wealth quartile range (2303 per 100,000) and remained very high for those in the lower middle and upper middle quartile ranges (1218 and 778 per 100,000, respectively). This is perhaps because the women, especially poor ones, may have sought help only once a serious complication had arisen. CONCLUSION: Achieving equitable coverage of all births by health professionals is still a distant goal in Indonesia, but even among women who receive professional care, maternal mortality ratios remain surprisingly high. This may reflect the limitations of home-based care. Phased introduction of fee exemption and transport incentives to enable all women to access skilled delivery care in health centres and emergency care in hospitals may be a feasible, sustainable way to reduce Indonesia's maternal mortality ratio.

Lay public's understanding of equipoise and randomisation in randomised controlled trials.

OBJECTIVES: To research the lay public's understanding of equipoise and randomisation in randomised controlled trials (RCTs) and to look at why information on this may not be not taken in or remembered, as well as the effects of providing information designed to overcome barriers. DESIGN: Investigations were informed by an update of systematic review on patients' understanding of consent information in clinical trials, and by relevant theory and evidence from experimental psychology. Nine investigations were conducted with nine participants. SETTING: Access (return to education), leisure and vocational courses at Further Education Colleges in the Midlands, UK. PARTICIPANTS: Healthy adults with a wide range of educational backgrounds and ages. INVESTIGATIONS: Participants read hypothetical scenarios and wrote brief answers to subsequent questions. Sub-samples of participants were interviewed individually to elaborate on their written answers. Participants' background assumptions concerning equipoise and randomisation were examined and ways of helping participants recognise the scientific benefits of randomisation were explored. MAIN OUTCOME MEASURES: Judgments on allocation methods; treatment preferences; the acceptability of random allocation; whether or not individual doctors could be completely unsure about the best treatment; whether or not doctors should reveal treatment preferences under conditions of collective equipoise; and how sure experts would be about the best treatment following random allocation vs doctor/patient choice. Assessments of understanding hypothetical trial information. RESULTS: Recent literature continues to report trial participants' failure to understand or remember information about randomisation and equipoise, despite the provision of clear and readable trial information leaflets. In current best practice, written trial information describes what will happen without offering accessible explanations. As a consequence, patients may create their own incorrect interpretations and consent or refusal may be inadequately informed. In six investigations, most participants identified which methods of allocation were random, but judged the random allocation methods to be unacceptable in a trial context; the mere description of a treatment as new was insufficient to engender a preference for it over a standard treatment; around half of the participants denied that a doctor could be completely unsure about the best treatment. A majority of participants judged it unacceptable for a doctor to suggest letting chance decide when uncertain of the best treatment, and, in the absence of a justification for random allocation, participants did not recognise scientific benefits of random allocation over normal treatment allocation methods. The pattern of results across three intervention studies suggests that merely supplementing written trial information with an explanation is unlikely to be helpful. However, when people manage to focus on the trial's aim of increasing knowledge (as opposed to making treatment decisions about individuals), and process an explanation actively, they may be helped to understand the scientific reasons for random allocation. CONCLUSIONS: This research was not carried out in real healthcare settings. However, participants who could correctly identify random allocation methods, yet judged random allocation unacceptable, doubted the possibility of individual equipoise and saw no scientific benefits of random allocation over doctor/patient choice, are unlikely to draw upon contrasting views if invited to enter a real clinical trial. This suggests that many potential trial participants may have difficulty understanding and remembering trial information that conforms to current best practice in its descriptions of randomisation and equipoise. Given the extent of the disparity between the assumptions underlying trial design and the assumptions held by the lay public, the solution is unlikely to be simple. Nevertheless, the results suggest that including an accessible explanation of the scientific benefits of randomisation may be beneficial provided potential participants are also enabled to reflect on the trial's aim of advancing knowledge, and to think actively about the information presented. Further areas for consideration include: the identification of effective combinations of written and oral information; helping participants to reflect on the aim of advancing knowledge; and an evidence-based approach to leaflet construction.

Randomisation in trials: do potential trial participants understand it and find it acceptable?

OBJECTIVE: To examine lay persons' ability to identify methods of random allocation and their acceptability of using methods of random allocation in a clinical trial context. DESIGN: Leaflets containing hypothetical medical, non-medical, and clinical trial scenarios involving random allocation, using material from guidelines for trial information leaflets. SETTING AND PARTICIPANTS: Adults attending further education colleges (n = 130), covering a wide range of ages, occupations, and levels of education. MAIN MEASURES: Judgements of whether each of five methods of allocation to two groups was random in a medical or non-medical scenario. Judgements of whether these allocation methods were acceptable in a randomised clinical trial scenario, with or without a scientific justification for randomisation. RESULTS: The majority of our group of participants judged correctly that allowing people their preference was not random, and that the following were random: using a computer with no information about the individual (recommended wording for MREC trial leaflets), tossing a coin, drawing a name out of a hat. Judgements were split over allocating people in turn (not a random allocation method but shares features with randomisation). Judgements were no different in medical and non-medical scenarios. Few of the correctly identified random methods were judged to be acceptable in a clinical trial scenario. Inclusion of a scientific justification for randomising significantly increased the acceptability of only one random method: allocation by computer. CONCLUSIONS: Current UK guidelines' recommended description of random allocation by computer seems warranted. However, while potential trial participants may understand what random allocation means, they may find it unacceptable unless offered an acceptable justification for its use.

Trials in surgery.

BACKGROUND: Trials in surgery pose some special problems. This paper examines these with reference to 10 years of methodological research sponsored by the UK National Health Service Research and Development programme. METHODS: Solutions to common problems encountered in surgical studies were considered, such as issues of blinding, dependence of results on technical skill and continued evolution of technology. RESULTS: Numerous methodological developments are described, including the tracker trial concept in which trial design can be adapted to take account of technical developments and interim results. The governance of trials, solutions to ethical conundra and the rising importance of databases are also discussed. CONCLUSION: Like surgery itself, the methodological toolkit for evaluation of surgical procedures continues to evolve. The rules of statistical and scientific probity provide plenty of scope for imaginative design solutions for surgical trials.

The measurement of active errors: methodological issues.

The value of research in any topic area turns on its validity. Patient safety research has revealed--or, at least, given renewed urgency to--a raft of methodological issues. The meaning and thus the value of empirical research in this field is contingent on getting the methodology right. The need for good methods for the measurement of error is necessary whenever an inference is intended and, since inferences lie at the heart of research and management, there is a huge need to understand better how to make measurements that are meaningful, precise, and accurate. In this paper we consider issues relating to the measurement of error and the need for more research.

Redesigning postnatal care: a randomised controlled trial of protocol-based midwifery-led care focused on individual women's physical and psychological health needs.

OBJECTIVES: To develop, implement and test the cost-effectiveness of redesigned postnatal care compared with current care on women's physical and psychological health. DESIGN: A cluster randomised controlled trial, with general practice as the unit of randomisation. Recruited women were followed up by postal questionnaire at 4 and 12 months postpartum and further data collected from midwife and general practice sources. SETTING: Thirty-six randomly selected general practice clusters in the West Midlands Health Region, UK. PARTICIPANTS: All women expected to be resident within recruited practices for postnatal care were eligible for inclusion. Attached midwives recruited 1087 women in the intervention and 977 in the control practice clusters. INTERVENTIONS: The systematic identification and management of women's health problems, led by midwives with general practitioner contact only when required. Symptom checklists and the Edinburgh Postnatal Depression Scale (EPDS) were used at various times to maximise the identification of problems, and individual care and visit plans based on needs. Evidence-based guidelines were used to manage needs. Care was delivered over a longer period. MAIN OUTCOME MEASURES: Women's health at 4 and 12 months, assessed by the Physical and Mental Component Scores (PCS and MCS) of the Short-Form 36 (SF-36) and the EPDS. Women's views about care, reported morbidity at 12 months, health service usage during the year, 'good practice' indicators and health professionals' views about care were secondary outcomes. RESULTS: At 4 and 12 months postpartum the mean MCS and EPDS scores were significantly better in the intervention group and the proportion of women with an EPDS score of 13+ (indicative of probable depression) was significantly lower relative to controls. The physical health score (PCS) did not differ. Health service usage was significantly less in the intervention group as well as reported psychological morbidity at 12 months. Women's views about care were either more positive or did not differ. Intervention midwives were more satisfied with redesigned care than control midwives were with standard care. Intervention care was cost-effective since outcomes were better and costs did not differ substantially. CONCLUSIONS: The redesigned community postnatal care led by midwives and delivered over a longer period, resulted in an improvement in women's mental health at 4 months postpartum, which persisted at 12 months and at equivalent overall cost. It is suggested that further research should focus on: the identification of postnatal depression through screening; whether fewer adverse longer term effects might be demonstrated among the children of the women who had the intervention care relative to the controls; testing interventions to reduce physical morbidity, including studies to validate measures of physical health in postpartum women. Further research is also required to investigate appropriate postnatal care for ethnic minority groups.

Top-rated British business research: has the emperor got any clothes?

Business schools have great prestige and charge large amounts of money for their courses. But how good is the science on which they base their prescriptions for action? To find out we examined the published output from the only three British business schools with the highest (5*) research assessment ranking at the time the articles were published. We conclude that theory development and model construction are often elegant. However, the methods used to obtain primary empirical information to confirm or refute the theories or populate models are poor, at least from a positivist or pragmatic ontological perspective. Large scale comparative studies made up only a small proportion of research output from the business schools. Literature reviews were not systematic. The sampling frame and rationale for selection of cases for study are inadequately described. The methods of data collection were frequently not given in sufficient detail to enable the study to be replicated and the conclusions tended to go far beyond what the data by themselves could support. However, this does not have to be the case-there are excellent examples of research in social sciences. We conclude, therefore, that top-rated British business research is a scantily clad emperor.

Helicobacter pylori infection in subjects with acute ischaemic stroke.

AIMS: To determine whether infection with Helicobacter pylori is a significant risk factor for stroke. SUBJECTS: A total 467 in-patients with clinical evidence of acute ischaemic stroke and 388 healthy controls with no evidence of cerebrovascular disease. METHODS: This was a case control study. The prevalence of Helicobacter pylori was measured by enzyme-linked immunosorbent assay in stroke patients and controls. A positive titre was defined as >15 U/ml and relationship with circulating plasma fibrinogen and social depravation was expressed using the Townsend Index. RESULTS: There were significantly more Helicobacter pylori positive individuals (274/398 (69%)) in the cases compared to the controls (206/352 (58.5%)). Fibrinogen levels were also significantly higher in Helicobacter pylori positive (mean 4.14, standard deviation 1.33) than negative individuals (mean 3.78, standard deviation 1.28). The association between Helicobacter pylori and stroke was lost in a logistic model controlling for socio-economic status. Furthermore, fibrinogen levels were not associated with Helicobacter pylori status in a linear regression model controlling for socio-economic status. CONCLUSIONS: Infection with Helicobacter pylori is associated with an increased risk of stroke and increased fibrinogen levels but these findings can be attributed to a confounding effect of socio-economic status.

The Leeds University Maternity Audit Project.

OBJECTIVES: To measure levels of and changes in compliance with evidence-based recommendations in obstetrics in the UK. To identify barriers to and factors associated with compliance. DESIGN: A quantitative case-note audit for 1988 and 1996, and a qualitative interview study of key staff. SETTING: Twenty maternity units, selected at random from all UK units SUBJECTS: Fifty consecutive cases of pre-term delivery (PTD), Caesarean section (CS), instrumental delivery (ID), and perineal repair (PR) operations in each period in each unit. The lead clinician, midwifery manager, a senior midwife, neonatologist, and middle-grade obstetrician in each unit. MAIN OUTCOME MEASURES: Maternal steroid use in PTD, antibiotic use in CS, use of the ventouse (vacuum extractor) rather than forceps as instrument of first choice for ID, and use of polyglycolic acid (PGA) sutures for PR in each time period. Facilities for implementing, staff attitudes to, and the degree of planning to follow each recommendation. MAIN RESULTS: The median proportion of ventouse as instrument of first choice in each unit was 8% (range 0-32%) in 1988, rising to 64% (range 0-98%) in 1996. PGA use for PR was 0% (range 0-30%) in 1988, and 72% (range 0-100%) in 1996. Steroid use for eligible PTD was median 0% (range 0-23%) in 1988, rising to 82% (range 63-95%) in 1996. Antibiotic use for CS was 7% (range 0-25%) rising to 84% (range 10-100%) in 1996. There was no relationship between unit size, type of unit, facilities, staff attitudes or degree of planning, and compliance with the recommendations, nor was the level of adherence to one standard typically correlated with adherence to the others. However, there was a positive correlation (R = 0.6, P < 0.005) between local availability of the Cochrane database of perinatal trials and unit compliance with the audit standards in the latter time period. CONCLUSIONS: We have documented a massive shift in practice in line with the evidence, although many units still have substantial room for improvement. About 2000 wound infections, 200 deaths due to prematurity, nearly 8000 women in pain from catgut sutures, and 1500 cases of severe perineal trauma from forceps remain preventable. The reasons why units vary remain obscure, although the qualitative interviews often revealed local factors such as key enthusiastic staff. There was no sign of evidence being positively driven into practice by any systematic managerial process. The relationship between Cochrane availability and high-standard care may be simply a marker of commitment to the evidence, but it remains plausible that if senior staff make Cochrane available for their juniors, audit compliance improves.

Effects of redesigned community postnatal care on womens' health 4 months after birth: a cluster randomised controlled trial.

BACKGROUND: Much postpartum physical and psychological morbidity is not addressed by present care, which tends to focus on routine examinations. We undertook a cluster randomised controlled trial to assess community postnatal care that has been redesigned to identify and manage individual needs. METHODS: We randomly allocated 36 general practice clusters from the West Midlands health region of the UK to intervention (n=17) or control (19) care. Midwives from the practices recruited women and provided care. 1087 (53%) of 2064 women were in practices randomly assigned to the intervention group, with 977 (47%) women in practices assigned to the control group. Care was led by midwives, with no routine contact with general practitioners, and was extended to 3 months. Midwives used symptom checklists and the Edinburgh postnatal depression scale (EPDS) to identify health needs and guidelines for the management of these needs. Primary outcomes at 4 months were obtained by postal questionnaire and included the women's short form 36 physical (PCS) and mental (MCS) component summary scores and the EPDS. Secondary outcomes were women's views about care. Multilevel analysis accounted for possible cluster effects. FINDINGS: 801 (77%) of 1087 women in the intervention group and 702 (76%) of 977 controls responded at 4 months. Women's mental health measures were significantly better in the intervention group (MCS, 3.03 [95% CI 1.53-4.52]; EPDS -1.92 [-2.55 to -1.29]; EPDS 13+ odds ratio 0.57 [0.43-0.76]) than in controls, but the physical health score did not differ. INTERPRETATION: Redesign of care so that it is midwife-led, flexible, and tailored to needs, could help to improve women's mental health and reduce probable depression at 4 months' postpartum.

Are randomized clinical trials good for us (in the short term)? Evidence for a "trial effect".

OBJECTIVE: To assess whether there is evidence that randomized controlled trials are systematically beneficial, or harmful, for patients. In other words, is there a "trial effect"? If so, to examine whether the evidence sheds light on the likely sources of the difference in outcomes. METHODS: Systematic review of the literature. RESULTS: We set out in some detail potential sources of a "trial effect" and potential biases. We found only 14 research articles (covering more than 21 trials) with relevant primary data. We extracted, with difficulty, quantitative data-sets from the articles, and classified these according to likely source of any apparent trial effect. The categories used were: differences in prognosis; superior treatment in the trial; and "protocol/Hawthorne effect" (benefit from improved routine care within a trial). ANALYSIS: The evidence available is limited in breadth (coming largely from cancer trials) and quality, as well as quantity. There is weak evidence to suggest that clinical trials have a positive effect on the outcome of participants. This does not appear to depend strongly on the trial demonstrating that an experimental treatment is superior. However, benefit to participants is less evident where scope for a "protocol/Hawthorne effect" was apparently limited (because there was no effective routine treatment or because the comparison group also received protocol care). A form of bias, arising if clinicians who tend to recruit to trials also tend to be better clinicians, could also explain these results. CONCLUSION: While the evidence is not conclusive, it is more likely that clinical trials have a positive rather than a negative effect on the outcome of patients. In the limited data available, the effect seems to be larger in trials where an effective treatment already exists and is included in the trial protocol. RECOMMENDATION: That carefully researched treatment protocols, and monitoring of outcomes, be used for all patients, not just those in trials.

Who's afraid of Thomas Bayes?

Sometimes direct evidence is so strong that a prescription for practice is decreed. Usually, things are not that simple-leaving aside the possibility that important trade offs may be involved, direct comparative data may be imprecise (especially in crucial sub-groups) or subject to possible bias, or there may be no direct comparative evidence; but still decisions have to be made. In these circumstances, indirect evidence-the plausibility of effects-enters the frame. But how should we describe the extent of plausibility and, having done so, how can this be integrated with any direct evidence that might exist. Also, how can allowance be made in a transparent (that is, explicit) way for perceptions of the size of bias in the direct evidence. Enter the Reverend Thomas Bayes; plausibility (however derived-laboratory experiment, qualitative study or just "experience") is captured numerically as degrees of belief ("prior" to the direct data) and updated (by the direct evidence) to yield "posterior" probabilities for use in decision making. The mathematical model used for this purpose must explicitly take account of assumptions about bias in the direct data. This paradigm bridges theory and practice, and provides the intellectual scaffold for those who recognise that (numerically definable) probabilities, and values (also numerically definable) underlie decisions, but who also realise that subjectivity is ineluctable in science.

The proportion of upper gastrointestinal symptoms in the community associated with Helicobacter pylori, lifestyle factors, and nonsteroidal anti-inflammatory drugs. Leeds HELP Study Group.

OBJECTIVE: Upper gastrointestinal disorders are common in the community, yet the determinants of these symptoms are poorly characterized. The association between upper gastrointestinal symptoms and Helicobacter pylori (H. pylori), socioeconomic status, nonsteroidal antiinflammatory drug (NSAID) use, smoking, alcohol, and coffee intake was assessed in a cross-sectional survey. METHODS: Subjects between the ages of 40-49 yr were randomly selected from the lists of 36 primary care centers. Participants attended their local primary care center and were interviewed by a researcher using a validated dyspepsia questionnaire. H. pylori status was determined by a nonfasting 13C-urea breath test. RESULTS: A total of 32,929 subjects were invited, and 8,407 (25%) attended and were eligible. Of these, 2,329 (28%) were H. pylori positive and 3,177 (38%) had dyspepsia. Also, 44% of H. pylori-infected participants reported dyspepsia compared with 36% of uninfected subjects [odds ratio = 1.39; 95% confidence interval (CI) 1.26-1.53]. H. pylori infection remained a significant risk factor for dyspepsia in a multiple logistic regression model (odds ratio = 1.21; 95% CI 1.09-1.34), suggesting that 5% of dyspepsia in the population is attributable to H. pylori. NSAIDs, low educational attainment, renting accommodation, absence of central heating, sharing a bed with siblings, and being married were also significantly associated with dyspepsia in this model. Smoking, but not drinking alcohol or coffee, was marginally associated with dyspepsia, but this finding was not robust. These factors were not associated with any dyspepsia subtype. CONCLUSIONS: H. pylori is significantly associated with dyspepsia and may be responsible for 5% of upper gastrointestinal symptoms in the community.

Can unequal be more fair? A response to Andrew Avins.

In this paper, we respond to Andrew Avins's recent review of methods whose use he advocates in clinical trials, to make them more ethical. He recommends in particular, "unbalanced randomisation". However, we argue that, before such a recommendation can be made, it is important to establish why unequal randomisation might offer ethical advantages over equal randomisation, other things being equal. It is important to make a pragmatic distinction between trials of treatments that are already routinely available and trials of restricted treatments. We conclude that unequal randomisation could, indeed, be an ethical compromise between protecting the interests of participants and those of society.

Clinical trials. A place for randomization in the interval between the end of recruitment and availability of results.

There is a time delay between the final recruitment of patients to a randomized controlled trial and the publication of results. The practical options available to decision makers during this gap can be listed according to whether all treatments are already widely available or whether at least one has been restricted to the trial. When the treatments are already in widespread use, the options are simply either to stop randomizing or to continue. When one trial treatment is restricted, there are further options: a) withdraw the restricted treatment altogether, pending the final analysis; b) continue to offer randomization, with a view to providing further data should these be needed; or c) make the intervention widely available to patients who would have previously been eligible for the trial. In this paper, we discuss the relative advantages and disadvantages of each option and discuss their attendant ethical implications. In particular, we suggest that continuing randomization is an option worthy of serious consideration. Randomizing patients acts as a "hedge" against the need for more data, given that sample size calculation is an inexact science. However, patients must be made aware of the basis on which randomization is offered.