The quality and diagnostic value of open narratives in verbal autopsy: a mixed-methods analysis of partnered interviews from Malawi.

BACKGROUND: Verbal autopsy (VA), the process of interviewing a deceased's family or caregiver about signs and symptoms leading up to death, employs tools that ask a series of closed questions and can include an open narrative where respondents give an unprompted account of events preceding death. The extent to which an individual interviewer, who generally does not interpret the data, affects the quality of this data, and therefore the assigned cause of death, is poorly documented. We aimed to examine inter-interviewer reliability of open narrative and closed question data gathered during VA interviews. METHODS: During the introduction of VA data collection, as part of a larger study in Mchinji district, Malawi, we conducted partner interviews whereby two interviewers independently recorded open narrative and closed questions during the same interview. Closed questions were collected using a smartphone application (mobile-InterVA) and open narratives using pen and paper. We used mixed methods of analysis to evaluate the differences between recorded responses to open narratives and closed questions, causes of death assigned, and additional information gathered by open narrative. RESULTS: Eighteen partner interviews were conducted, with complete data for 11 pairs. Comparing closed questions between interviewers, the median number of differences was 1 (IQR: 0.5-3.5) of an average 65 answered; mean inter-interviewer concordance was 92% (IQR: 92-99%). Discrepancies in open narratives were summarized in five categories: demographics, history and care-seeking, diagnoses and symptoms, treatment and cultural. Most discrepancies were seen in the reporting of diagnoses and symptoms (e.g., malaria diagnosis); only one pair demonstrated no clear differences. The average number of clinical symptoms reported was 9 in open narratives and 20 in the closed questions. Open narratives contained additional information on health seeking and social issues surrounding deaths, which closed questions did not gather. CONCLUSIONS: The information gleaned during open narratives was subject to inter-interviewer variability and contained a limited number of symptom indicators, suggesting that their use for assigning cause of death is questionable. However, they contained rich information on care-seeking, healthcare provision and social factors in the lead-up to death, which may be a valuable source of information for promoting accountable health services.

Identifying humanitarian crises in population surveillance field sites: simple procedures and ethical imperatives.

OBJECTIVES: Effective early warning systems of humanitarian crises may help to avert substantial increases in mortality and morbidity, and prevent major population movements. The Butajira Rural Health Programme (BRHP) in Ethiopia has maintained a programme of epidemiological surveillance since 1987. Inspection of the BRHP data revealed large peaks of mortality in 1998 and 1999, well in excess of the normally observed year-to-year variation. Further investigation and enquiry revealed that these peaks related to a measles epidemic, and a serious episode of drought and consequent food insecurity that went undetected by the BRHP. This paper applies international humanitarian crisis threshold definitions to the BRHP data in an attempt to identify suitable mortality thresholds that may be used for the prospective detection of humanitarian crises in population surveillance sites in developing countries. STUDY DESIGN: Empirical investigation using secondary analysis of longitudinal population-based cohort data. METHODS: The daily, weekly and monthly thresholds for crises in Butajira were applied to mortality data for the 5-year period incorporating the crisis periods of 1998-1999. Days, weeks and months in which mortality exceeded each threshold level were identified. Each threshold level was assessed in terms of prospectively identifying the true crisis periods in a timely manner whilst avoiding false alarms. RESULTS: The daily threshold definition is too sensitive to accurately detect impending or real crises in the population surveillance setting of the BRHP. However, the weekly threshold level is useful in identifying important increases in mortality in a timely manner without the excessive sensitivity of the daily threshold. The weekly threshold level detects the crisis periods approximately 2 weeks before the monthly threshold level. CONCLUSION: Mortality measures are highly specific indicators of the health status of populations, and simple procedures can be used to apply international crisis threshold definitions in population surveillance settings for the prospective detection of important changes in mortality rate. Standards for the timely use of surveillance data and ethical responsibilities of those responsible for the data should be made explicit to improve the public health functioning of current sentinel surveillance methodologies.

The Distribution and Effects of Child Mortality Risk Factors in Ethiopia: a Comparison of Estimates from DSS and DHS

Objectives: To conduct a comparative analysis of the distribution and effects of under-five mortality correlates using Demographic and Health Survey (DHS) and Demographic Surveillance System (DSS) data from Ethiopia; and to investigate the methodological bias in DHS-based childhood mortality rates due to the impossibility of including children whose mothers were deceased. Methods: Using all-cause under-5 mortality as an outcome variable; the distribution and effects of risk factors were modeled using survival analysis. All live births in rural Ethiopia in the 5-year period before the 2005 DSS+ survey and between 01/01/2000 and 31/12/2004 in the DSS in the Butajira Rural Health Program (in the Southern Nations; Nationalities; and People's (SNNP) region of Ethiopia) were included. Results: Overall; similar estimates of hazard rate ratios were derived from both DHS and DSS data and the child mortality risk profile is similar between each data source; with multiple births and living in less populous households being significant risk factors for under-five mortality. Nevertheless; some notable differences were observed. The DSS data was more sensitive to local variations in population composition and health status; whilst the more dispersed DHS approach tended to average out local variation across the country. Excluding children whose mothers were deceased from the DSS analysis had no important effect on risk profiles or estimates of survival functions at age 5 years. DHS survival functions were somewhat lower than DSS estimates (BRHP=0.87; DHS rural Ethiopia=0.67; DHS SNNP=0.66). Conclusion: Despite differing methodologies; cross-sectional DHS and longitudinal DSS data produce estimates of the distribution and effects of under-five mortality risk factors that are broadly similar. The differing methodological characteristics of DHS and DSS mean that when combined; these two data sources have the potential to provide a comprehensive picture of national population composition and health status as well as the extent of local variation - both of which are important for health monitoring and planning

Violent behaviour by psychiatric patients.

There is no substantial body of evidence to prove that psychiatric patients as a group are more violent than the general population. In determining whether violence is exhibited, the underlying personality of the individual is far more important than the psychiatric illness from which he suffers. Individuals with psychopathic personalities, those who abuse or are dependent on alcohol drugs, and those with pronounced paranoid or sadistic traits may be characterized by aggressive and violent behaviour. There is at present no reliable method of predicting if or when an individual patient will exhibit assaultive behaviour. Psychiatric hospitals, especially in the UK, are characterized more by frequent petty violence than by serious violence. Victims of assaults by patients are often related or known to them an interact with them on a personal level; they may also play a considerable role consciously or otherwise in instigating the assault.

A study of violent behaviour among patients in psychiatric hospitals.

This study of violent behaviour among psychiatric in-patients in three hospitals found that many incidents of petty violence occurred and serious violence was rare. Among the offending patients there were more females than males, more young than old and the most commonly represented diagnosis was schizophrenia. Nursing staff were by far the commonest victims. Most of the incidents occurred during the daytime and were perpetrated mainly by patients nursed on acute psychiatric wards. The vast majority of patients were nonviolent.

"Asset stripping" the declining mental hospital.

A psychogeriatric day hospital was developed from a long-stay back ward of a large mental hospital whose population of patients had decreased by half. Women in the 70-80 age group with varying degrees of dementia benefited most from such facilities. In view of the current cut-back in Government spending one of the greatest assets in providing a comprehensive service for the elderly mentally infirm may well be found in the declining mental hospital.

Long-stay patients with long-stay drugs. A case review; a cause for concern.

The psychotropic medication of 200 long-stay psychiatric patients was reviewed by the consutant in charge of the ward, the ward doctor, and the ward sister. Those patients in whom medication was altered or stopped were reassessed by the same team after 6 months. About half the patients on medication had been receiving unnecessary or excessive medication.

Identification and definition of long-stay mental hospital population.

In a study to identify and define a central group of long-stay psychiatric patients who are resistant to discharge 422 were found who had been in hospital for eight to 23 years. These "problem" patients represented 40% of all long-stay patients. The total number of years that they had spent in hospital was nearly double that of the remaining patients. They were mostly unvisited, unoccupied, and single, and most were suffering from schizophrenia or organic psychosis. Half were in a good or reasonably good state of mental health, and three-quarters were in a good or reasonably good state of physical health. The findings have implications for the rehabilitation and treatment of these patients and also for the provision of community aftercare facilities.