Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative.

AIM: To improve the accuracy and completeness of reporting of studies of diagnostic accuracy in order to allow readers to assess the potential for bias in a study and to evaluate the general isability of its results. METHODS: The standards for reporting of diagnostic accuracy (STARD) steering committee searched the literature to identify publications on the appropriate conduct and reporting of diagnostic studies and extracted potential items into an extensive list. Researchers, editors, and members of professional organisations shortened this list during a 2 day consensus meeting with the goal of developing a checklist and a generic flow diagram for studies of diagnostic accuracy. RESULTS: The search for published guidelines about diagnostic research yielded 33 previously published checklists, from which we extracted a list of 75 potential items. At the consensus meeting, participants shortened the list to a 25-item checklist, by using evidence whenever available. A prototype of a flow diagram provides information about the method of recruitment of patients, the order of test execution and the numbers of patients undergoing the test under evaluation, the reference standard, or both. CONCLUSIONS: Evaluation of research depends on complete and accurate reporting. If medical journals adopt the checklist and the flow diagram, the quality of reporting of studies of diagnostic accuracy should improve to the advantage of clinicians, researchers, reviewers, journals, and the public.

[Reporting studies of diagnostic accuracy according to a standard method; the Standards for Reporting of Diagnostic Accuracy (STARD)]. / Verslaglegging van diagnostisch evaluatieonderzoek volgens een standaardmethode; de 'standards for reporting of diagnostic accuracy' (STARD).

The objective of the 'Standards for Reporting of Diagnostic Accuracy' (STARD) initiative is to improve the reporting of studies of diagnostic accuracy, so as to allow readers to assess the potential for bias in a study and to evaluate the generalibility of its results. The group searched the literature to identify publications on the appropriate conduct and reporting of diagnostic studies. This was used to draw up a list of potential items. During a consensus meeting, a group of researchers, medical journal editors, and members of professional organisations reduced this list to a usable checklist. Wherever possible, evidence from the literature was used to justify the decisions made. The search for published guidelines about diagnostic research yielded 33 previously published checklists, from which a list of 75 potential items was extracted. At the consensus meeting, participants shortened the list to a 25-item checklist. A generic flow diagram was drawn up to provide guidance on the method for including patients, the order in which tests were to be conducted and the number of patients to undergo the test being evaluated, the reference standard, or both. A scientific publication can only be assessed when the reporting is both correct and complete. Use of the checklist and flow diagram will improve the quality of reports produced, to the advantage of clinicians, researchers, reviewers, journal editors and other interested parties.

Evidence-based medicine: useful tools for decision making.

Evidence-based medicine (EBM) integrates clinical experience and patient values with the best available research information. There are four steps in incorporating the best available research evidence in decision making: asking answerable questions; accessing the best information; appraising the information for validity and relevance; and applying the information to patient care. Applying EBM to individual patients requires drawing up a balance sheet of benefits and harms based on research and individual patient data. The most realistic and efficient use of EBM by clinicians at the point of care involves accessing and applying valid and relevant summaries of research evidence (evidence-based guidelines and systematic reviews). The future holds promise for improved primary research, better EBM summaries, greater access to these summaries, and better implementation systems for evidence-based practice. Computer-assisted decision support tools for clinicians facilitate integration of individual patient data with the best available research data.

Telling people about screening programmes and screening test results: how can we do it better?

To make an informed choice about whether to be screened, people need information that allows them to weigh up the benefits and harms of screening. To understand their screening test results they require even more information. Yet currently, people attending a screening programme or considering a screening test may only be told that the test can detect disease or risk factors for disease, and that early intervention improves outcomes. When given their test results, people are generally only told the test was abnormal ("positive") or normal ("negative"). We believe that information given before and after the screening test can, and should, be improved. This will probably require information that includes both the benefits and harms of screening and is probabilistic. Indeed, we believe the traditional dichotomisation of screening test results into positive and negative is problematic, and could be replaced by standard use of risks or probabilistic data before and after screening. The relevant risk data could be explained in a range of ways, for example, quantitatively, qualitatively, and/or by "anchoring" to everyday experiences. In this paper we explore why dichotomisation of screening test results is problematic and look at the adverse consequences of presenting test results in terms of true and false, positive and negative. We present some ideas on alternative ways of providing information on screening programmes and screening test results. Our aim is to stimulate debate about these issues and to provide some starting points which could be further developed and evaluated in a wide range of screening programmes.

Screening for Type 2 diabetes mellitus: a decision analytic approach.

AIMS: Screening for asymptomatic Type 2 diabetes mellitus has been advocated on the grounds that diabetes is a common condition associated with increased morbidity and mortality, but uncertainty remains about the impact of early treatment. This study aimed to determine whether the potential benefits of screening are likely to outweigh the potential harm and to explore which variables significantly influence the balance of benefit and harm resulting from screening. METHODS: A decision analysis comparing the relative impact of using a single fasting blood glucose screening test, between the ages of 45 and 60 years, with the impact of not screening. The model weighs the increase in quality adjusted life years (QALYs) from reduction in microvascular and cardiovascular complications against the potential decrease in QALYs associated with earlier diagnosis and treatment in an asymptomatic population. RESULTS: The baseline model suggests a saving of 10 QALYs for every 10,000 individuals screened: a gain of four from postponed microvascular complications and 17 from avoided cardiovascular complications, as opposed to a loss of 11 as a result of earlier diagnosis in screening detected cases. The balance of benefit and harm is sensitive to baseline cardiovascular risk, the effectiveness of cardiovascular interventions and the relative disutility assigned to early diagnosis and treatment for an individual without symptoms. CONCLUSIONS: The immediate disutility of earlier diagnosis and additional treatment may be greater than the potential long-term benefit from postponing microvascular complications. Screening decisions should therefore be based largely on consideration of cardiovascular risk and the availability of evidence based interventions to reduce cardiovascular risk.

Does treatment of vesicoureteric reflux in childhood prevent end-stage renal disease attributable to reflux nephropathy?

OBJECTIVE: It is believed that end-stage renal disease (ESRD) attributable to reflux nephropathy is preventable by the active treatment of vesicoureteric reflux in childhood with long-term antibiotics and ureteric reimplantation surgery. We aimed to test this belief. METHODOLOGY: The Australia and New Zealand Dialysis and Transplant Registry of new patients 5 to 44 years of age treated for ESRD between 1971 and 1998, categorized by age and primary renal disease, was used to analyze the age-specific incidences of ESRD attributable to reflux nephropathy using a before-after study design. The early 1960s were regarded as the introduction period for the active treatment of childhood vesicoureteric reflux. A time-delay in treatment effect was expected. Patients with ESRD attributable to other causes were used as a comparative group. RESULTS: The incidence of ESRD attributable to reflux nephropathy and nonreflux nephropathy has increased. For reflux nephropathy, the rate of change was significantly associated with age, with a downward trend in incidence with decreasing age suggesting a minor treatment effect. This trend was no longer evident when adjustment was made for changing diagnostic practices. An opposite trend was observed for the nonreflux nephropathy group, who demonstrated an upward trend in incidence with decreasing age. CONCLUSIONS: Treatment of children with vesicoureteric reflux has not been accompanied by the hoped-for reduction in the incidence of ESRD attributable to reflux nephropathy. A randomized trial with a control (no-treatment) arm is required to appropriately assess the medical belief that long-term antibiotics and surgery improve the natural history of vesicoureteric reflux.

A new, clinically-based algorithm for the diagnosis of HIV in African tuberculosis patients: cross-sectional analysis from Mzuzu, Malawi.

We aimed to create an improved, clinically-based algorithm for the diagnosis of HIV in tuberculosis (TB) patients. Cross-sectional analysis was performed on data from adult TB patients consecutively diagnosed at a Malawian district level hospital. Of 225 patients, 187 with valid HIV results were included in the study. Sixty-seven per cent were HIV seropositive. Urban address, history of skin rash and sexually transmitted diseases (STDs) and, on examination, oral candidiasis and lymphadenopathy, were associated with HIV co-infection. Using these clinical characteristics, a case definition for HIV was constructed. The Mzuzu clinical case definition was highly sensitive (86%). The area under the receiver operating characteristic (ROC) curve was 0.81, significantly larger than existing World Health Organization (WHO) clinical case definitions. The Mzuzu definition is proposed for further evaluation in settings where HIV serological testing is not readily available.

Variability in the interpretation of dimercaptosuccinic acid scintigraphy after urinary tract infection in children.

UNLABELLED: Technetium-99m-dimercaptosuccinic acid (DMSA) scintigraphy is a frequently used diagnostic test in pediatric practice to assess the presence and severity of renal damage. Most commonly it is performed after urinary tract infection. The aim of this study was to investigate the variability in the interpretation of DMSA scans by pediatric nuclear medicine physicians in this clinical setting. METHODS: We selected all 441 scans from children with first-time urinary tract infection who presented between 1993 and 1995 to a pediatric casualty department and who are participants in a prospective cohort study. Two hundred and ninety-four scans were performed at a median time of 7 days after diagnosis, and 147 scans were from children who were free from further infection over a 1-yr follow-up period. Two experienced nuclear medicine physicians independently interpreted the 441 scans according to whether renal damage was present or absent and using the modified 4-level grading system for DMSA abnormality of Goldraich. Apart from being informed that urinary tract infection was the indication for DMSA scintigraphy, no other clinical information was given to the nuclear medicine physicians. The indices of variability used were the percentage of agreement and the kappa statistic. For the grading scale used, both measures were weighted with integers representing the number of categories from perfect agreement. Disagreement was analyzed for children, kidneys and kidney zones. RESULTS: There was agreement in 86% (kappa = 69%) for the normal-abnormal DMSA scan dichotomy, and the weighted agreement was 94% (weighted kappa = 82%) for the grading of abnormality. Disagreement of DMSA scan interpretation of > or =2 grades was present in three cases (0.7%). The same high level of agreement was present for patient, kidney and kidney zone comparisons. Agreement was not influenced by age or timing of scintigraphy after urinary tract infection. CONCLUSION: Two experienced nuclear medicine physicians showed good agreement in the interpretation of DMSA scintigraphy in children after urinary tract infection and using the grading system of Goldraich.

Symptomatic urinary tract infection in preschool Australian children.

OBJECTIVE: To describe the demographic and clinical features, short-term outcomes, microbiology and renal tract abnormalities of a cohort of young Australian children with symptomatic urinary tract infection. METHODOLOGY: A total of 304 children < 5 years with their first documented symptomatic urinary tract infection who presented consecutively to the Emergency Department of a paediatric hospital between March 1993 and December 1994 and without a known predisposing cause were identified and details of their acute illness were recorded. Renal tract sonography, micturating cystourethrography and Tc-99 m dimercaptosuccinic acid scintigraphy (DMSA) were routinely performed. RESULTS: Of those who presented with urinary tract infection, 169 were boys and 135 girls; 64% were less than 1 year of age. For children from the local community, the cumulative incidence of urinary tract infection within the first 5 years of life was estimated to be 1.9% for boys and 1.8% for girls. There were no significant differences in illness characteristics according to mode of referral or geographical locality. Presenting symptoms were generally nonspecific and not referrable to the urinary tract. There were no deaths. One per cent of children required ventilatory support, and bacteraemia occurred in 6%, all of whom were under 6 months of age. E. coli was the causal organism in 84%, and a high in vitro resistance to ampicillin/ amoxycillin (54%) was demonstrated by the pathogens isolated. Bacteriuria was eradicated in 99% with antimicrobial treatment. In this setting, the sensitivities of dipstick urinalysis (leucocyte esterase+/-nitrites) and pyuria on microscopy (>10 x 10(6) white cells L(-1)) were 85%. Abnormal DMSA scintigraphy was detected in 39%, vesicoureteric reflux in 28%, and obstructive uropathy in 1%. CONCLUSIONS: This study provides current and local data on a large sample of children <5 years with urinary tract infection, which are useful to clinicians who manage children at risk of the condition.

Use of systematic reviews of randomised trials by Australian neonatologists and obstetricians.

OBJECTIVE: To determine what proportion of Australian neonatologists and obstetricians report using systematic reviews of randomised trials. DESIGN: Cross-sectional survey using structured telephone interviews. SETTING: Australian clinical practice in 1995. PARTICIPANTS: 103 of the 104 neonatologists in Australia (defined as clinicians holding a position in a neonatal intensive care unit); a random sample of 145 members of the Royal Australian College of Obstetricians and Gynaecologists currently practising in Australia. MAIN OUTCOME MEASURES: Information sources used in clinical practice; reported awareness of, access to and use of systematic reviews, and consequent practice changes. RESULTS: Response rates were 95% (neonatologists) and 87% (obstetricians); 71 neonatologists (72%) and 55 obstetricians (44%) reported using systematic reviews, primarily for individual patient care. Databases of systematic reviews were used with a median frequency of once per month. Among neonatologists, systematic reviews were used more commonly by those who were familiar with computers, attended professional meetings, and had authored research papers. Among obstetricians, they were used more commonly by those who were familiar with computers, had less than 10 years' clinical experience, attended more deliveries, and were full-time staff specialists in public hospitals. Of neonatologists who reported using systematic reviews, 58% attributed some practice change to this use. For obstetricians, the corresponding figure was 80%. CONCLUSIONS: There is evidence that Australian neonatologists and obstetricians use systematic reviews and modify their practice accordingly. Dissemination efforts can benefit from knowledge of factors that predict use of systematic reviews.

Trends in the health burden due to urinary tract infection in children in Australia.

OBJECTIVE: To estimate the health burden of urinary tract infection in children less than 15 years of age in Australia and to ascertain whether any significant change has occurred during the past decade. METHODOLOGY: The number of children less than 15 years of age who were admitted in New South Wales for urinary tract infection between 1981 and 1994 was ascertained from the Department of Health, and age and sex specific incidence rates were calculated using Australian Bureau of Statistics population data. Costs for inpatient care were calculated using the cost weights from Australia National Disease Related Groups Version 3 for urinary tract infection (DRG 577). The frequency of the four most commonly requested renal tract imaging procedures in children following urinary tract infection and which qualified for Medicare reimbursement were obtained from the Health Insurance Commission for 1984-1994: micturating cystourethrography, intravenous urography, renal ultrasonography, and nuclear medicine renal studies. RESULTS: There were 1203 children who were admitted with urinary tract infection in New South Wales in 1994, at an estimated cost of $A1.6 million. Since 1981, the age standardized annual incidence of urinary tract infection requiring hospitalization has increased from 0.5 to 0.9 per 1000 children, largely because of an increase in the number of young children admitted (from 0.6 to 2.0 per 1000 children less than 5 years of age). In 1994, 46,230 non-inpatient renal imaging procedures were undertaken in children under 15 years of age at a cost of $A5.3 million. CONCLUSIONS: Urinary tract infection is an important and increasing health problem for Australian children, particularly for preschool children. Whether this represents a true increase in the incidence of urinary tract infection or improved diagnosis and more intensive management is not possible to establish with this study design. Prospective population based studies are required to assess more completely the frequency with which urinary tract infection occurs in children and any changes that may be occurring.

Variation in the diagnosis of vesicoureteric reflux using micturating cystourethrography.

Variability in the interpretation of micturating cystourethrography by paediatric radiologists for the diagnosis of vesicoureteric reflux in children was evaluated. All 265 micturating cystourethrograms (MCUs) that were available from 304 consecutive children aged 0.5-61 months-who were investigated after their first urine infection between 1993 and 1995 as part of a prospective cohort study-were selected for interpretation. Three experienced paediatric radiologists from the same department independently interpreted the MCUs according to the grading system of the International Reflux Study in Children, from grades 0 to V, with the presence of intrarenal reflux also noted. Apart from being informed that urine infection was the indication for the MCU, no other clinical information was given to the radiologists. The indices of variability used were the percentage of agreement and the kappa statistic, expressed as a percentage. Both measures were weighted with integers representing the number of categories from perfect agreement. Disagreement was analysed for children and kidneys. For the diagnosis of vesicoureteric reflux in individual patients, including grade, the percentage of agreement was 96%-97% (kappa 90%-91%) and the weighted percentage of agreement was 96%-98% (weighted kappa 93%-94%). The same high level of agreement was present for individual kidneys, with a percentage of agreement of 97%-98% (kappa 89%-92%) and a weighted percentage of agreement of 98%-99% (kappa 94%-95%). There was near perfect agreement in the interpretation of radiological micturating cystourethrography among three experienced paediatric radiologists for the diagnosis and grade of vesicoureteric reflux. Any variations in the medical care of children suspected of having vesicoureteric reflux are not explained by differences in the reporting of this diagnostic test.

Do commercial serological kits for Helicobacter pylori infection differ in accuracy? A meta-analysis.

OBJECTIVES: To compare the accuracy of common commercial serological kits for Helicobacter pylori and to ascertain factors affecting accuracy. METHODS: A comprehensive MEDLINE and manual search strategy was used to identify all articles comparing two or more kits. Each article was critically appraised for sample characteristics, study design, and data handling. The data comparing accuracy of the kits was analyzed by standard statistical methods as well as summary receiver operator characteristic curves (sROCs). A sROC also was used to estimate overall test accuracy and to identify factors affecting the measurement of accuracy. RESULTS: The 21 studies identified were of varying quality, but our analyses suggested that different commercial kits did not have significantly different accuracy. Overall, at a sensitivity of 85%, specificity was estimated to be 79%. Test accuracy measured was significantly higher in studies with smaller proportions of infected patients. CONCLUSIONS: There is little evidence in the literature to suggest that any one of the common commercial serological kits is more accurate than any other. The overall accuracy of these kits may not be adequate for clinical decision-making in all patient groups.

Predicting intention to use condoms in homosexual men: an application and extension of the theory of reasoned action.

PIP: A questionnaire based upon the Theory of Reasoned Action was developed and tested to identify predictors of intention to use condoms. Behavioral intentions, attitudes and subjective normative beliefs, behavioral norms, and age of HIV-positive status were included on the questionnaire. Internal consistency of the four components was high, with Cronbach's alpha coefficients of 0.76-0.87. Qualitative data were collected over 2 months from 194 homosexual men who participated in Sydney's homosexual community. Participants aged 17-64 years of mean age 29.5 attended individual interviews or small focus groups to discuss their views concerning condom use. 15% reported having tested HIV-positive, 59% tested negative, and 26% reported not knowing their HIV serostatus. According to logistic modeling, the significant predictors of intentions to use a condom were attitudes and behavioral norms, while HIV antibody positive status and age directly influence behavioral intentions to use a condom. About 50% of men over age 25 years intended to use a condom, irrespective of HIV status. Overall, 59% of men aged 25 years and younger intended to use a condom; 22% of HIV-positive men and 63% of HIV-negative men.^ieng

An evidence based approach to individualising treatment.

To which groups of patients can the results of clinical trials be applied? This question is often inappropriately answered by reference to the trial entry criteria. Instead, the benefit and harm (adverse events, discomfort of treatment, etc) of treatment could be assessed separately for individual patients. Patients at greatest risk of a disease will have the greatest net benefit as benefit to patients usually increases with risk while harm remains comparatively fixed. To assess net benefit, the relative risks should come from (a meta-analysis of) randomised trials; the risk in individual patients should come from multivariate risk equations derived from cohort studies. However, before making firm conclusions, the assumptions of fixed adverse effects and constant reduction in relative risk need to be checked.

Fourier transform infrared spectroscopy of dysplastic, papillomavirus-positive cervicovaginal lavage specimens.

To assess the utility of a new, rapid, economical procedure that may prove valuable in cervical screening, Fourier transform infrared (ir) spectroscopy was performed on 25 cervicovaginal lavage specimens from women referred for colposcopy on the basis of a cytological abnormality detected on their Pap smear and whose lavage specimen was positive for human papillomavirus. Of the 18 classed as CIN I or less by histopathology, 11 showed band frequencies that deviated only slightly from spectra that characterize normal cervical cells and 3 of 5 "atypia" specimens had spectra identical to normal. Two of 3 classed as CIN II had spectra only slightly more abnormal to these 11. In the case of 2 graded as CIN I, several bands were similarly altered in the direction of the pattern seen for 4 CIN III specimens. A further CIN I sample gave a spectrum that was even further shifted toward the latter and the remaining CIN I sample had a pattern that matched the 4 CIN IIIs. The most obvious change in each of the CIN IIIs was an additional peak at 972 cm-1 and this has been suggested as a key indicator for malignancy. One of the 3 CIN IIs had this peak. Other characteristic spectral changes were seen as well in the CIN III samples. High-risk HPV18 was present in 3 of the CIN III samples, as well as in one specimen classed as atypia, but having an abnormal ir spectrum. Low-risk HPV 6 or 11 was seen along in samples with a normal or slightly abnormal ir spectrum, but never in those that showed an ir pattern that was abnormal. The current study has therefore shown complete concordance between ir spectral findings and histopathology result in the case of CIN III specimens, but less precise matching for other grades of CIN. The spectral differences revealed by ir spectroscopy are likely to characterize molecular abnormalities in cervical cells during progression to cancer and may therefore have potential in assisting with clinical decision making. More studies will, however, be required to establish the place of this technique in cervical screening.

Who has Pap smears in New South Wales? Patterns of screening across sociodemographic groups.

The aim of this study was to determine whether Pap smear screening at adequate intervals is associated with area of residence, frequency of consultations with a general practitioner, socioeconomic status and non-English-speaking background. A representative 10 per cent sample of women from New South Wales and the Australian Capital Territory, aged 25 to 69 years and registered with the Health Insurance Commission (Medicare) (N = 155,281) was used to obtain age, postcode, frequency of Pap smears and frequency of consultations with general practitioners in the three-year period from February 1985 to January 1988. Census data for each postcode area were used as an indicator of other socio-demographic characteristics. Age-specific screening rates did not vary between Sydney, Newcastle/Wollongong, Canberra, and nonmetropolitan areas. In all age groups, having had a smear was most strongly associated with the frequency with which a woman consulted a general practitioner. Women who visited a general practitioner at least four times a year on average were about twice as likely to have had a recent Pap smear as those who averaged less than one visit per year. Screening rates were lowest among women living in areas with the most non-English-speakers and the lowest socio-economic status. Sociodemographic factors and health service usage patterns influence the proportion of women who are currently being screened. Evaluation of interventions to improve Pap smear screening rates should consider whether the percentage of women screened increases overall, and also whether the imbalances in screening rates between different groups are diminishing.