Patient safety in developing countries: retrospective estimation of scale and nature of harm to patients in hospital.

OBJECTIVE: To assess the frequency and nature of adverse events to patients in selected hospitals in developing or transitional economies. DESIGN: Retrospective medical record review of hospital admissions during 2005 in eight countries. SETTING: Ministries of Health of Egypt, Jordan, Kenya, Morocco, Tunisia, Sudan, South Africa and Yemen; the World Health Organisation (WHO) Eastern Mediterranean and African Regions (EMRO and AFRO), and WHO Patient Safety. PARTICIPANTS: Convenience sample of 26 hospitals from which 15,548 patient records were randomly sampled. MAIN OUTCOME MEASURES: Two stage screening. Initial screening based on 18 explicit criteria. Records that screened positive were then reviewed by a senior physician for determination of adverse event, its preventability, and the resulting disability. RESULTS: Of the 15,548 records reviewed, 8.2% showed at least one adverse event, with a range of 2.5% to 18.4% per country. Of these events, 83% were judged to be preventable, while about 30% were associated with death of the patient. About 34% adverse events were from therapeutic errors in relatively non-complex clinical situations. Inadequate training and supervision of clinical staff or the failure to follow policies or protocols contributed to most events. CONCLUSIONS: Unsafe patient care represents a serious and considerable danger to patients in the hospitals that were studied, and hence should be a high priority public health problem. Many other developing and transitional economies will probably share similar rates of harm and similar contributory factors. The convenience sampling of hospitals might limit the interpretation of results, but the identified adverse event rates show an estimate that should stimulate and facilitate the urgent institution of appropriate remedial action and also to trigger more research. Prevention of these adverse events will be complex and involves improving basic clinical processes and does not simply depend on the provision of more resources.

Factors associated with haemoglobin concentration among Timor-Leste children aged 6-59 months.

The study was conducted to assess the prevalence of and factors associated with haemoglobin (Hb) concentrations among children aged 6-59 months in Timor-Leste. The 2003 Demographic and Health Survey was a multi-stage cluster survey of 4,320 households from four different geographic regions in Timor-Leste. In total, 4,514 children aged 6-59 months were included in the analysis. The prevalence of anaemia (Hb concentration <11.0g/dL) was 38.2% (638/1,668) for children aged 6-23 months and 22.6% (644/2,846) for older children (p<0.001). Girls had a higher mean Hb concentration than boys (11.9g/dL vs 11.7g/ dL, p<0.006) and children who had diarrhoea in the previous two weeks had a lower Hb concentration than children without diarrhoea (11.5g/dL vs 11.9g/dL, p<0.001). Children from the richest and middle-class households had a lower average Hb concentration than those from the poorest households (11.8g/ dL, 11.7g/dL vs 12.0g/dL, p<0.001). Children of mothers with some secondary or more education had a lower mean Hb concentration than children of mothers with completed primary, some primary and no education (11.7 g/dL vs 11.9 g/dL, 11.8 g/dL, and 11.9 g/dL, p=0.002). Children from severely-anaemic mothers had a lower mean Hb concentration than children from moderately-, mild and not anaemic mothers (10.5 g/dL vs 11.1 g/dL, 11.6 g/dL, 12.0 g/dL, p<0.001). After backward stepwise hierarchical multiple regression, wasting, male sex, recent diarrhoea, household wealth index (richest and middle-class), maternal educational status (some secondary or more and some primary), and maternal anaemic status were significantly associated with a lower Hb concentration in children and increased age of child and duration of breastfeeding (6 months) with a higher Hb concentration. Anaemia-prevention programmes among children in Timor-Leste should focus on those children aged less than two years, children with recent diarrhoea, wasted children, high socioeconomic status, and anaemic mothers.

Analysing differences in clinical outcomes between hospitals.

OBJECTIVE: To examine the variation between hospitals in rates of severe intraventricular haemorrhage (IVH) in preterm babies adjusting for case mix and sampling variability. DESIGN: Cross sectional study of pooled data from 1995 to 1997. SETTING: 24 neonatal intensive care units (NICUs) in the Australian and New Zealand Neonatal Network. PARTICIPANTS: 5413 infants of gestational age 24-30 weeks. MAIN OUTCOME MEASURES: Crude rates of severe (grades 3 and 4) IVH and rates adjusted for case mix using logistic regression, and for sampling variability using shrinkage estimators. RESULTS: The overall rate of severe IVH was 6.8%, but crude rates for individual units ranged from 2.9 to 21.4%, with interquartile range (IQR) 5.7-8.1%. Adjusting for the five significant predictor variables--gestational age at birth, 1 minute Apgar score, antenatal corticosteroids, transfer after birth, and sex--actually increased the variability in rates (IQR 5.9-9.7%). Shrinkage estimators, which adjust for differences in unit sizes and outcome rates, reduced the variation in rates (IQR 6.3-7.5%). Adjusting for case mix and using shrinkage estimators showed that one unit had a significantly higher adjusted rate than expected, while another was significantly lower. If all units could achieve an average rate equal to the 20th centile (5.74%), then 60 cases of severe IVH could be prevented in a 3 year period. CONCLUSIONS: The use of shrinkage estimators may have a greater impact on the variation in outcomes between hospitals than adjusting for case mix. Greater reductions in morbidity may be achieved by concentrating on the best rather than the worst performing hospitals.

Can the Expanded Disability Status Scale be assessed by telephone?

Information from patients who are unable to continue their visits to a study centre may be of major importance for the interpretation of results in multiple sclerosis (MS) clinical trials. To validate a questionnaire based on the Expanded Disability Status Scale (EDSS), patients in five different European centres were assessed independently by pairs of trained EDSS raters, first by telephone interview and a few days later by standardized neurological examination. Seventy women and 40 men with an average age of 43.7 years (range 19-74 years) were included in the study. Mean EDSS score at the last visit was 4.5 (0-9). EDSS assessment by telephone was highly correlated with the EDSS determined by physical examination (Pearson's correlation coefficient = 0.95). An intraclass correlation coefficient (ICC) of 94.8% was found for the total sample; 77.6% and 86%, respectively, for patients with EDSS < 4.5 (n = 46) and > 4.5 (n = 64). Kappa values for full agreement were 0.48; for variation by +0.5 steps and +1.0 steps, 0.79 and 0.90, respectively. Best agreement could be found in higher EDSS scores, where assessment by telephone interview might be needed most. The telephone questionnaire is a valid tool to assess EDSS score in cases where the patient is unable to continue visiting a study centre or in long-term follow-up of trial participants.

Adverse events in surgical patients in Australia.

OBJECTIVE: To determine the adverse event (AE) rate for surgical patients in Australia. DESIGN: A two-stage retrospective medical record review was conducted to determine the occurrence of AEs in hospital admissions. Medical records were screened for 18 criteria and positive records were reviewed by two medical officers using a structured questionnaire. SETTING: Admissions in 1992 to 28 randomly selected hospitals in Australia. STUDY PARTICIPANTS: Five hundred and twenty eligible admissions were randomly selected from in-patient database in each hospital. A total of 14,179 medical records were reviewed, with 8747 medical and 5432 surgical admissions. MAIN OUTCOME MEASURES: Measures included the rate of AEs in surgical and medical admissions, the proportion resulting in permanent disability and death, the proportion determined to be highly preventable, and the identification of risk factors associated with AEs. RESULTS: The AE rate for surgical admissions was 21.9%. Disability that was resolved within 12 months occurred in 83%, 13% had permanent disability, and 4% resulted in death. Reviewers found that 48% of AEs were highly preventable. The risk of an AE depended on the procedure and increased with age and length of stay. CONCLUSION: The high AE rate for surgical procedures supports the need for monitoring and intervention strategies. The 18 screening criteria provide a tool to identify admissions with a greater risk of a surgical AE. Risk factors for an AE were age and procedure, and these should be assessed prior to surgery. Prophylactic interventions for infection and deep vein thrombosis could reduce the occurrence of AEs in hospitals.

Effect of a community action program on adult quit smoking rates in rural australian towns: the CART project.

BACKGROUND: This article describes one outcome of a randomized controlled trial of community action for cancer prevention. The aims of this article were to (a) explore the effectiveness of a community action program in decreasing community smoking rates in rural Australian towns and (b) describe the relationship between adult smoking quit and uptake rates and demographic variables. METHODS: In 1992, 20 towns were selected for randomization. Community action involved formation of community committees and utilization of access point networks to initiate and maintain intervention strategies. At post-test, outcomes were proportion of "quitters" from a cohort of self-described smokers, proportion of "uptakers" from a cohort of self-described nonsmokers, and "net effect." RESULTS: Differences in quit rate, uptake rate, and net effect for intervention compared to control condition favored the intervention in all cases, although mainly nonsignificant. Significantly more male smokers quit in intervention towns than in control towns [7.0% (95% CI: 0.6, 13.5)]. CONCLUSIONS: Given that CART utilized and improved upon strategies argued as effective in the literature, the limited success of the project in reducing adult smoking, considered in combination with COMMIT findings, suggests the need for further innovation in the field.

A comparison of iatrogenic injury studies in Australia and the USA. I: Context, methods, casemix, population, patient and hospital characteristics.

OBJECTIVE: To better understand the differences between two iatrogenic injury studies of hospitalized patients in 1992 which used ostensibly similar methods and similar sample sizes, but had quite different findings. The Quality in Australian Health Care Study (QAHCS) reported that 16.6% of admissions were associated with adverse events (AE), whereas the Utah, Colorado Study (UTCOS) reported a rate of 2.9%. SETTING: Hospitalized patients in Australia and the USA. DESIGN: Investigators from both studies compared methods and characteristics and identified differences. QAHCS data were then analysed using UTCOS methods. MAIN OUTCOME MEASURES: Differences between the studies and the comparative AE rates when these had been accounted for. RESULTS: Both studies used a two-stage chart review process (screening nurse review followed by confirmatory physician review) to detect AEs; five important methodological differences were found: (i) QAHCS nurse reviewers referred records that documented any link to a previous admission, whereas UTCOS imposed age-related time constraints; (ii) QAHCS used a lower confidence threshold for defining medical causation; (iii) QAHCS used two physician reviewers, whereas UTCOS used one; (iv) QAHCS counted all AEs associated with an index admission whereas UTCOS counted only those determining the annual incidence; and (v) QAHCS included some types of events not included in UTCOS. When the QAHCS data were analysed using UTCOS methods, the comparative rates became 10.6% and 3.2%, respectively. CONCLUSIONS: Five methodological differences accounted for some of the discrepancy between the two studies. Two explanations for the remaining three-fold disparity are that quality of care was worse in Australia and that medical record content and/or reviewer behaviour was different.

Analysis of dichotomous outcome data for community intervention studies.

Community intervention trials are becoming increasingly popular as a tool for evaluating the effectiveness of health education and intervention strategies. Typically, units such as households, schools, towns, counties, are randomized to receive either intervention or control, then outcomes are measured on individuals within each of the units of randomization. It is well recognized that the design and analysis of such studies must account for the clustering of subjects within the units of randomization. Furthermore, there are usually both subject level and cluster level covariates that must be considered in the modelling process. While suitable methods are available for continuous outcomes, data analysis is more complicated when dichotomous outcomes are measured on each subject. This paper will compare and contrast several of the available methods that can be applied in such settings, including random effects models, generalized estimating equations and methods based on the calculation of 'design effects', as implemented in the computer package SUDAAN. For completeness, the paper will also compare these methods of analysis with more simplistic approaches based on the summary statistics. All the methods will be applied to a case study based on an adolescent anti-smoking intervention in Australia. The paper concludes with some general discussion and recommendations for routine design and analysis.

Survival of patients with colorectal cancer detected by a community screening program.

OBJECTIVE: To determine survival rates for people with colorectal cancer detected through Bowelscan, a community screening program. DESIGN: Survey of data from local medical practitioners, and comparison with data from State cancer registries. SUBJECTS AND SETTING: 249 people with colorectal cancer detected after faecal occult blood screening in north-eastern New South Wales, 1987-1996. Follow-up was in 1998-1999. MAIN OUTCOME MEASURES: Five-year survival rates and relative survival ratios. RESULTS: Five-year survival rates for the screen-detected cancer patients were 90% for those with Dukes' stage A cancers, 75% for Dukes' B, 52% for Dukes' C and 0 for Dukes' D (although one person with Dukes' D cancer was living at four-year follow-up at the end of the study). Because of the higher percentage of Dukes' A cases in the population whose cancer was detected through screening, the resulting five-year relative survival ratio was significantly better than for those recorded by New South Wales, South Australian and Queensland cancer registries: 0.82 (95% confidence interval, 0.74-0.90) compared to 0.59 (P < or = 0.001). CONCLUSIONS: The study supports the findings of three overseas randomised trials that screening reduces mortality from colorectal cancer. We estimate that screening 200,000 people would detect about 250 colorectal cancers and prevent as many as 55 deaths.

Using clinical indicators to identify areas for quality improvement.

Clinical indicators (CI) are increasingly being used to assess the quality of health care being provided by physicians and hospitals. However, a standardised reporting format and a methodology to assess the utility of the CI data has not been developed. This paper provides the reporting format that has been developed for the clinical colleges. The results for four surgical indicators are used to illustrate how the CI data can determine the potential to improve the quality of care. Numerical estimates of the potential gains that could be made are calculated by: (i) determining the outcome if the current mean rate was shifted to the rate for the best 20% of hospitals and (ii) identifying units with unusual variation in rates and shifting their rate to the average. All four indicators reveal gains that could impact on health policy and clinical practice.

An analysis of the causes of adverse events from the Quality in Australian Health Care Study.

OBJECTIVE: To examine the causes of adverse events (AEs) resulting from healthcare to assist in developing strategies to minimise preventable patient injury. DESIGN: Descriptions of the 2353 AEs previously reported by the Quality in Australian Health Care Study (QAHCS) were reviewed. A qualitative approach was used to develop categories for human error and for prevention strategies to minimise these errors. These categories were then used to classify the AEs identified in the QAHCS, and the results were analysed with previously reported preventability and outcome data. RESULTS: 34.6% of the causes of AEs were categorised as "a complication of, or the failure in, the technical performance of an indicated procedure or operation", 15.8% as "the failure to synthesise, decide and/or act on available information", 11.8% as "the failure to request or arrange an investigation, procedure or consultation", and 10.9% as "a lack of care and attention or failure to attend the patient". AEs in which the cause was cognitive failure were associated with higher preventability scores than those involving technical performance. The main prevention strategies identified were "new, better, or better implemented policies or protocols" (23.7% of strategies), "more or better formal quality monitoring or assurance processes" (21.2%), "better education and training" (19.2%), and "more consultation with other specialists or peers" (10.2%). CONCLUSION: The causes of AEs or errors leading to AEs can be characterised, and human error is a prominent cause. Our study emphasises the need for designing safer systems for care which protect the patient from the inevitability of human error. These systems should provide new policies and protocols and technological support to aid the cognitive activities of clinicians.

New measures for reporting the magnitude of small-area variation in rates.

Measuring the variation in health outcomes, for example, mortality, morbidity, hospitalization, across small areas is an accepted way of screening large amounts of routinely-collected data. Although simple measures of variation, for example, the extremal quotient, are intuitively appealing, they have poor statistical properties. More sophisticated measures, based on hierarchical models, have better statistical properties, but are in a form that is foreign to most public health officials. The analyses in this paper converted the small-area variance obtained from a hierarchical model into three new measures: the ratio of high versus low rates across small areas, and the percentage and number of adverse events, such as deaths, that might be avoidable if the causes of the variation between areas could be removed. The approach was applied to mortality data from New South Wales, Australia. The three new measures can help public health officials make judgements about whether to proceed with more detailed (and expensive) studies without having to rely on the statistical significance of an obscure index.

Community action for health promotion: a review of methods and outcomes 1990-1995.

OBJECTIVE: Our objective was to explore the effectiveness of community action as a strategy for health promotion, through a critical review of five years of community action evaluation literature. METHODS: Community action was defined as a health promotion, program that involved the community in implementation and control of the process of the program. Criteria for scientific evaluation of programs were proposed for sampling and control procedures, reliability and validity of instruments, analysis techniques, and specification of details of the intervention. A critical review of the literature, located by an on-line and related reference search, was undertaken for community action aimed at reducing cancer and cardiovascular disease, between January 1990 and May 1995. RESULTS: None of seven community action studies (17 articles) that examined cancer risk factors fulfilled all the criteria for rigorous scientific evaluation. The most methodologically adequate cancer study, the COMMIT intervention, had only a moderate degree of success in reducing community smoking rates. Similarly, none of the six studies (25 articles) on cardiovascular disease fulfilled all the criteria. The results for the most methodologically adequate study, the Minnesota Heart Health Program, were disappointing, with strong secular trends preventing adequate assessment of the intervention effect. CONCLUSIONS: The finding that none of the reviewed studies met all evaluation criteria was due to several factors, including political considerations, feasibility, and the continued evolution of the science of evaluation in health promotion. Some important questions are posed for researchers by the failure of methodologically superior projects, such as COMMIT, to show major gains in reducing health risk behaviors.

Estimating potential savings in cancer deaths by eliminating regional and social class variation in cancer survival in the Nordic countries.

STUDY OBJECTIVES: To examine equity in the health care system with regard to cancer patient care by estimating the level of systematic regional variation in cancer survival in the Nordic countries. Specifically, those cancer sites which exhibit high levels of systematic regional variation in survival and hence inequity were identified. Estimating the reduction in cancer deaths which could be achieved by eliminating this variation so that everyone receives effective care will provide a readily interpretable measure of the amount of systematic regional variation. A comprehensive analysis of regional variation in survival has not previously been conducted so appropriate statistical methodology must be developed. SETTING AND PARTICIPANTS: All those aged 0-90 years who had been diagnosed with at least one of 12 common malignant neoplasms between 1977 and 1992 in Denmark, Finland, Norway, and Sweden. DESIGN: A separate analysis was conducted for each country. Regression models for the relative survival ratio were used to estimate the relative risk of excess mortality attributable to cancer in each region after correcting for age and sex. An estimate of the amount of systematic regional variation in survival was obtained by subtracting the estimated expected random variation from the observed regional variation. An estimate was then made of the potential reduction in the number of cancer deaths for 2008-12 if regional variation in survival were eliminated so that everyone received the same level of effective care. MAIN RESULTS: Between 2008 and 2012, an estimated 2.5% of deaths from cancers in the 12 sites studied could be prevented by eliminating regional variation in survival. The percentage of potentially avoidable deaths did not depend on country or sex but it did depend on cancer site. There was no relationship between the level of regional variation in a given country and the level of survival. The cancer sites for which the greatest percentage savings could be achieved were melanoma (11%) and cervix uteri (6%). The sites for which the highest number of deaths could be prevented were prostate, colon, melanoma, and breast. CONCLUSIONS: This methodology showed a small amount of systematic regional variation in cancer survival in the Nordic countries. The cancer sites with high levels of regional variation identified are potential targets for cancer control programmes.

The annual variation in activity and funding for acute public hospitals in NSW, 1988-89 to 1992-93.

Casemix-based funding was introduced into the Victorian health system without an assessment of the annual variation in inpatient activity. Before undertaking such a funding reform, it would be appropriate to determine the level of annual variation in inpatient activity for individual hospitals that could be attributable to chance or random variation. If the annual random variation is not accounted for, then casemix-based funding may actually lead to inefficiencies. For this study, hospital inpatient activity and funding data for 120 acute public hospitals from New South Wales for the years 1988-89 to 1992-93 were used to estimate the standard deviation of the annual random variation in activity and gross operating payment. Through linear regression, estimates of the standard deviation of random variation about the underlying trend were obtained for each hospital. The results showed that, depending on the size of the hospital, total diagnosis related group cost weights have a standard deviation in the range of 2 to 16 per cent of total activity, whilst gross operating payment has an equivalent standard deviation that ranges from 1 to 10 per cent annually. The magnitude of the variation would suggest that funding of hospitals should either be based on average activity over several years or based on bands of activity in order to reduce the potential random variation in funding levels.

Monitoring postoperative pulmonary embolisms in Australia--I. Variation between hospitals.

Information on inter-hospital variation in clinical outcomes can be extracted from routinely collected morbidity data in Australia. Postoperative pulmonary embolism rates, one of seven clinical performance indicators now subject to review during Australian hospitals' quality accreditation surveys, is used as an example throughout. Using indirect statistical standardisation and Poisson models of random variation, we were able to identify five hospitals which recorded statistically significantly higher pulmonary embolism rates than average and three which recorded lower than expected rates. The application of these methods to all hospital outcome measures will assist in objectively monitoring the quality of patient care.

Monitoring postoperative pulmonary embolisms in Australia--II. Variation within hospitals.

Clinical indicator data can be analysed quickly and efficiently within hospitals to provide quality review staff with monthly or quarterly reports on their own hospital's performance relative to national or regional normative statistics. The statistical tools used are indirect standardisation to correct for casemix differences between institutions and control charts, and cumulative sum charts for the analysis and presentation of findings. Routinely collected postoperative pulmonary embolism data are used to illustrate these approaches.

The Quality in Australian Health Care Study.

A review of the medical records of over 14,000 admissions to 28 hospitals in New South Wales and South Australia revealed that 16.6% of these admissions were associated with an "adverse event", which resulted in disability or a longer hospital stay for the patient and was caused by health care management; 51% of the adverse events were considered preventable. In 77.1% the disability had resolved within 12 months, but in 13.7% the disability was permanent and in 4.9% the patient died.