Search PubMed⌕ Search

Biomedical subjects

Tom Marshall

Publications and source records attributed to Tom Marshall.

33 records · Page 2Linked to original sources
PubMed2004

Are visual measures of mood superior to questionnaire measures in non-Western settings?

OBJECTIVE: We hypothesised that, in a non-Western setting where literacy was not universal, a visual measure (the FACES test) would be more valid than a traditional psychiatric questionnaire [the General Health Questionnaire (GHQ)] as a screening test for mood disorders. METHODS: The study was nested within a randomised controlled trial of 450 patients with a common mental disorder (CMD). Subjects were evaluated at 2, 6 and 12 months after enrolment with the Clinical Interview Standardised-Revised (CISR) (the gold standard), the GHQ5 (the shortened version of the GHQ-12) and the FACES test. RESULTS: Correlation coefficients and Receiver Operating Characteristic (ROC) curves show superiority of the GHQ5 in the detection of CMD based on the CISR, compared with the visual FACES test. Less-educated subjects had particular difficulty completing the FACES. The kappa coefficient of agreement between the two instruments, using the cut-off point for the GHQ5 estimated by the ROC curves, was between 0.70 and 0.75. CONCLUSIONS: Contrary to our hypothesis, the questionnaire-based measure was significantly superior to a visual measure of mood, especially for less-educated subjects. A short five-item version of the GHQ has a good discriminatory ability for CMD and may be used as a brief alternative to standardised interviews in clinical and survey settings.

Adolescent↗
PubMed2004

A randomized controlled trial of league tables and control charts as aids to health service decision-making.

OBJECTIVES: Health service managers and other decision-makers are required to act on the basis of data. Little attention has been paid to the effects of data presentation on the decisions taken. This study uses a randomized controlled trial design to investigate the effects of two forms of data presentation--league tables and control charts--on health service decision-makers. METHODS: Directors of public health in 122 health authorities in the UK were mailed three case studies and a questionnaire. The case studies showed data on variations in mortality by health service provider. The questionnaire asked them to indicate whether they would take action as a result of the data and to identify the health service providers on whom they would take action. Participants were randomly allocated to receive the same data in the form of ranked histograms (league tables) or control charts. MAIN OUTCOME MEASURE: The percentage of participants who would take action on health service providers. RESULTS: Fifty-seven questionnaires were returned. For each case study, respondents receiving data as league tables stated they would take action on significantly more health service providers than those receiving data as control charts: for the first case study, the percentages were 3.3% versus 1.8% (P < 0.001) for league tables and control charts, respectively; for the second case study, 15.9% versus 6.7% (P = 0.029), respectively; and for the third, 5.9% versus 0.7% (P = 0.002), respectively. Respondents receiving data as league tables were significantly more likely to request further information on case mix. CONCLUSIONS: Compared with league tables, health service decision-makers identify fewer outliers for further action when performance data are presented as control charts. They also reduce the tendency to request further information. Using control charts rather than league tables for the routine presentation of comparative data would reduce over-investigation of unusual performance.

Bias↗
PubMed2004

Counting in-hospital deaths in England: a comparison of hospital computer systems and mortuary registers.

OBJECTIVES: In-hospital death counts derived from hospital computer systems have been used in England by an independent company, 'Dr Foster', to rank the quality of care of hospitals, but the validity of the underlying data remains unclear. This study compares counts of in-hospital deaths using two different sources - the hospital computer system and the mortuary register - to determine: whether the counts of in-hospital deaths from these two sources differed; qualitative explanations for possible sources of discrepancy; the direction and magnitude of any differences; and the possible impact of any differences on the Dr Foster rankings. METHODS: The four highest and the four lowest National Health Service (NHS) hospitals in the West Midlands, as ranked by Dr Foster, participated. Each hospital was asked to compare the monthly counts of in-hospital deaths from the hospital computer system and the hospital mortuary register for the fiscal year 1999/2000. RESULTS: One hospital, with a computerised mortuary register, had identical counts of in-hospital deaths. Two hospitals reported 4-5% more deaths and four hospitals reported fewer deaths (0.4-7%) from their hospital computer system than from their mortuary register. These differences were not large enough to change their Dr Foster rankings. Wide discrepancies were noted on a monthly basis (range: -13.9% to +15.9%). DISCUSSION: The differences between the two sources of in-hospital death counts were not large enough to influence the Dr Foster ranks but were sufficient to raise concern about the validity and completeness of mortality data in the NHS.

Attitude of Health Personnel↗
PubMed2003

Coronary heart disease prevention: insights from modelling incremental cost effectiveness.

OBJECTIVE: To determine which treatments for preventing coronary heart disease should be offered to which patients by assessing their incremental cost effectiveness. DESIGN: Modelling study. DATA SOURCES: Cost estimates (for NHS) and estimates of effectiveness obtained for aspirin, antihypertensive drugs, statins and clopidogrel. DATA SYNTHESIS: Treatment effects were assumed to be independent, and cost per coronary event prevented was calculated for treatments individually and in combination across patients at a range of coronary risks. RESULTS: The most cost effective preventive treatments are aspirin, initial antihypertensive treatment (bendrofluazide and atenolol), and intensive antihypertensive treatment (bendrofluazide, atenolol and enalapril), whereas simvastatin and clopidogrel are the least cost effective (cost per coronary event prevented in a patient at 10% coronary risk over five years is 3500 pounds sterling for aspirin, 12 500 pounds sterling for initial antihypertensives, 18 300 pounds sterling for intensive antihypertensives, 60 000 pounds sterling for clopidogrel, and 61 400 pounds sterling for simvastatin). Aspirin in a patient at 5% five year coronary risk costs less than a fifth as much per event prevented (7900 pounds sterling) as simvastatin in a patient at 30% five year risk (40 800 pounds sterling). DISCUSSION: A cost effective prevention strategy would offer aspirin and initial antihypertensive treatment to all patients at greater than 7.5% five year coronary risk before offering statins or clopidogrel to patients at greater than 15% five year coronary risk. Incremental cost effectiveness analysis of treatments produces robust, practical cost effectiveness rankings that can be used to inform treatment guidelines.

Antihypertensive Agents↗
PubMed2003

Understanding variation in quality improvement: the treatment of sore throats in primary care.

BACKGROUND: In 1988, two practices attempted to improve the prescribing of antibiotics for sore throat. The initiative produced only modest improvements in prescribing practice, a finding the authors found difficult to explain. This paper reanalyses the data from an audit of antibiotic prescribing for sore throat in general practice. OBJECTIVE: Our aim was to demonstrate the use of Shewhart control charts and to obtain fresh insight into the variations in clinical practice revealed in clinical audit data. METHODS: We use Shewhart control charts to explore variation in antibiotic prescribing between GPs and to suggest the action most likely to result in improvement. RESULTS: Using control charts, it is possible to distinguish two categories of GPs: low prescribers of antibiotics and high prescribers of antibiotics. Low prescribers of antibiotics show common cause variation, indicating that their prescribing is a stable process. Among low prescribers, improvement can best be achieved by changing the common underlying process. One high prescriber of antibiotics is affected by special cause variation. Among high prescribers, improvement can best be achieved by investigating the special causes affecting this GP and learning lessons from the findings. CONCLUSION: The original improvement effort took the same action on all GPs in both practices. Our analysis suggests that such an approach was unlikely to be successful and that different actions were needed for high and low prescribers. The control charts provide fresh insights on the original data and guide improvement efforts.

Anti-Bacterial Agents↗
PubMed2002

Resource implications and health benefits of primary prevention strategies for cardiovascular disease in people aged 30 to 74: mathematical modelling study.

OBJECTIVE: To develop a model to determine resource costs and health benefits of implementing guidelines for the prevention of cardiovascular disease in primary care. DESIGN: Modelling of data from six strategies for prevention of cardiovascular disease. Strategies incorporated two ways of identifying patients for assessment: traditional (assessment of all adults) and novel (preselection of patients for assessment using a prior estimate of their risk of cardiovascular disease). Three treatment strategies were modelled in conjunction with each identification strategy. SETTING: England. SUBJECTS: Patients aged 30 to 74 eligible for primary prevention strategies for cardiovascular disease who were selected from a hypothetical population of 2000. MAIN OUTCOME MEASURES: Resource costs of assessing eligible adults, providing treatment and follow up to those eligible, and number of cardiovascular events this should prevent. RESULTS: Novel strategies prevented more cardiovascular disease, at lower cost, than traditional strategies. Some treatment strategies prevent more cardiovascular disease with fewer resources than others. The findings were robust across a range of different assumptions about workload. CONCLUSION: Preselecting patients for assessment makes better use of staff time than assessing all adults. Treating many patients with low cost drugs is more efficient than prescribing a few patients intensive antihypertensives and statins. Authors of guidelines should model workload implications and health benefits of following their recommendations.

Adult↗
PubMed2002

Number of boys born to men exposed to polychlorinated byphenyls.

We studied the sex of children born to individuals involved in the Yucheng oil disaster, Taiwan, who were exposed to polychlorinated byphenyls (PCBs) after an oil contamination accident in 1979. Men exposed to PCBs before age 20 years had a lower chance of having a baby boy than did age-matched and neighbourhood-matched controls (odds ratio 0.65, 95% CI 0.45-0.93). The male-to-female sex ratio of children born to men exposed to PCBs after age 20 years, however, approached that seen in controls (0.90, 0.59-1.35). We noted no significant difference in the birth ratio of infants born to exposed and unexposed mothers (0.93, 0.77-1.12). Our findings suggest that paternal exposure to PCBs before age 20 years affects the sex of a subsequently born child.

Adult↗
PubMed2002

Understanding variation for clinical governance: an illustration using the diagnosis and treatment of sore throat.

BACKGROUND: The aim of clinical governance is to improve clinical care. An understanding of the information contained in variation is central to any improvement effort. We must distinguish between variation intrinsic to a process (common cause variation) and variation caused by extrinsic factors (special cause variation). The control chart is a method of distinguishing between these two kinds of variation: it is used in industry to effect improvement and may be useful in primary care. AIM: To illustrate the use of control charts to distinguish between common cause and special cause variation and to guide appropriate action. DESIGN OF STUDY: Analysis of diagnostic and treatment decisions for sore throat. SETTING: Single practice in the West Midlands. METHODS: We identified each general practitioner's (GP's) consultations for sore throat over a two-year period. We grouped these into two diagnostic categories (tonsillitis and non-tonsillar throat infection) and two treatment categories (antibiotics and no antibiotics). These data were illustrated graphically as XY control charts. RESULTS: In this practice, a special cause affects one GP's diagnosis--he is less likely to use the term 'tonsillitis'. A special cause also affects his treatment decisions--he is more likely to prescribe antibiotics. Diagnostic and treatment differences between the remaining GPs are consistent with common cause variation. CONCLUSION: In this practice, action to improve the quality of diagnosis and treatment of sore throat shouldfocus on investigating why one practitioner's diagnosis and treatment differs from that of his colleagues. Control chart analysis is valuable because it enables users to obtain practical guidance for action.

Decision Making↗
PubMedDate not supplied

Misleading measurements: modeling the effects of blood pressure misclassification in a United States population.

OBJECTIVE: The clinical diagnosis of hypertension is subject to misclassification, and this may be clinically important. This article calculates positive and negative predictive values for blood pressure measurement and assesses the frequency of clinically important blood pressure misclassification. DESIGN, SETTING, AND PARTICIPANTS: A modeling study was carried out on 4763 adults in the National Health and Nutrition Examination Survey (NHANES) POPULATION: True treatment eligibility was determined by applying Joint National Committee (JNC) VII criteria to individuals in the study population. Each individual was also allocated a series of blood pressures incorporating an error term reflecting day-to-day measurement variation. Test positives are persons classified as needing treatment on the basis of the mean of 2 blood pressure measurements. MEASUREMENTS AND MAIN RESULTS: Positive predictive values of a diagnosis of hypertension based on 2 measurements were calculated for each age-sex group. Low-risk false positives and highrisk false negatives were categorized as clinically important errors. Positive predictive values are high in persons older than age 65. In persons ages 16 to 34, the positive predictive value is 0.24 (95% confidence interval [CI]: 0.17-0.32) in men and 0.16 (95% CI: 0.06-0.26) in women. Persons younger than age 35 are almost always at low risk of cardiovascular disease, and therefore this misclassification is clinically important. Even with 24-hour ambulatory blood pressure measurement, positive predictive values in young adults are under 0.5. CONCLUSIONS: Blood pressure estimation is a poor diagnostic test in low-prevalence populations such as young adults. Estimation of blood pressure should be informed by prior estimation of cardiovascular risk.

Adolescent↗