Community Water Improvement, Household Water Insecurity, and Women's Psychological Distress: An Intervention and Control Study in Ethiopia.

BACKGROUND: Over 650 million people worldwide lack access to safe water supplies, and even among those who have gained access to 'improved' sources, water may be seasonally unreliable, far from homes, expensive, and provide insufficient quantity. Measurement of water access at the level of communities and households remains crude, and better measures of household water insecurity are urgently needed to inform needs assessments and monitoring and evaluation. We set out to assess the validity of a quantitative scale of household water insecurity, and to investigate (1) whether improvements to community water supply reduce water insecurity, (2) whether water interventions affect women's psychological distress, and (3) the impacts of water insecurity on psychological distress, independent of socio-economic status, food security, and harvest quality. METHODS AND FINDINGS: Measures were taken before and one to six months after a community water supply improvement in three villages in rural northern Ethiopia. Villages similar in size and access to water sources and other amenities did not receive interventions, and served as controls. Household water insecurity was assessed using a 21-item scale based on prior qualitative work in Ethiopia. Women's psychological distress was assessed using the WHO Self-Reporting Questionnaire (SRQ-20). Respondents were either female heads of household or wives of the heads of household (n = 247 at baseline, n = 223 at endline); 123 households provided data at both rounds. The intervention was associated with a decline of approximately 2 points on the water insecurity scale between baseline and endline compared to the control (beta -1.99; 95% CI's -3.15, -0.84). We did not find evidence of impact of the intervention on women's psychological distress. Water insecurity was, however, predictive of psychological distress (p <0.01), independent of household food security and the quality of the previous year's harvest. CONCLUSION: These results contribute to the construct validity of our water insecurity scale, and establish our approach to measuring water insecurity as a plausible means of evaluating water interventions. Improvements to community water supplies were effective in reducing household water insecurity, but not psychological distress, in this population. Water insecurity was an important predictor of psychological distress. This study contributes to an emerging literature on quantitative assessment of household water insecurity, and draws attention to the potential impact of improved access to water on women's mental well-being.

Assessment of beta-cell function during the oral glucose tolerance test by a minimal model of insulin secretion.

OBJECTIVE: To characterise the performance of beta-cell during a standard oral glucose tolerance test (OGTT). DESIGN: Fifty-six subjects were studied. A minimal analogic model of beta-cell secretion during the OGTT was applied to all OGTTs (see below). The amount of insulin secreted over 120' in response to oral glucose (OGTT-ISR; Insulin Units 120'-1 m-2 BSA) and an index of beta-cell secretory 'force' (beta-Index; pmol.min-2.m-2 BSA) were computed with the aid of the model. In protocol A, 10 healthy subjects underwent two repeat 75 g OGTT with frequent (every 10'-15') blood sampling for glucose and C-peptide to test the reproducibility of OGTT-ISR and beta-Index with a complete or a reduced data set. In protocol B, 7 healthy subjects underwent three OGTTs (50, 100 or 150 g), to test the stability of the beta-Index under different glucose loads. In protocol C, 29 subjects (15 with normal glucose tolerance, 7 with impaired glucose tolerance and 7 with newly diagnosed type 2 diabetes) underwent two repeat 75 g OGTT with reduced (every 30' for 120') blood sampling to compare the reproducibility and the discriminant ratio (DR) of OGTT-ISR and beta-index with the insulinogenic index (IG-Index: Delta Insulin 30' - Basal/Delta Glucose 30' - Basal). In protocol D, 20 subjects (14 with normal glucose tolerance, 5 with impaired glucose tolerance and 1 with newly-diagnosed type 2 diabetes) underwent a 75 g OGTT and an intravenous glucose tolerance test (IVGTT) on separate days to explore the relationships between acute (0'-10') insulin response (AIR) during the IVGTT and beta-index and OGTT-ISR during the OGTT. RESULTS: In all protocols, the minimal analogic model of C-peptide secretion achieved a reasonable fit of the experimental data. In protocol A, a good reproducibility of both beta-index and OGTT-ISR was observed with both complete and reduced (every 30') data sets. In protocol B, increasing the oral glucose load caused progressive increases in OGTT-ISR (from 2.63 +/- 0.70 to 5.11 +/- 0.91 Units.120'-1.m-2 BSA; P < 0.01), but the beta-index stayed the same (4.14 +/- 0.35 vs. 4.29 +/- 0.30 vs. 4.30 +/- 0.33 pmol.min-2.m-2 BSA). In protocol C, both OGTT-ISR and beta-index had lower day-to-day CVs (17.6 +/- 2.2 and 12.4 +/- 2.4%, respectively) and higher DRs (2.57 and 1.74, respectively) than the IG-index (CV: 35.5 +/- 6.3%; DR: 0.934). OGTT-ISR was positively correlated to BMI (P < 0.03), whereas beta-index was inversely related to both fasting and 2 h plasma glucose (P < 0.01 for both). In protocol D, beta-index, but not OGTT-ISR, was significantly correlated to AIR (r = 0.542, P < 0.02). CONCLUSIONS: Analogically modelling beta-cell function during the OGTT provides a simple, useful tool for the physiological assessment of beta-cell function.

Comparison of simulated patient and didactic methods of teaching HIV risk assessment to medical residents.

INTRODUCTION: Physicians need to be well trained in HIV risk assessment interview skills. Little has been written in the literature concerning training methods for this specialized interview. METHODS: One model to teach the HIV risk assessment interview has been developed and has been used to teach third-year medical students. We compared this interactive model, which uses simulated patients to teach HIV Risk Assessment, to a didactic one. Twelve medical residents were taken through either the interactive session or the didactic session. Pre-post changes from questionnaires were calculated to determine any differences in sessions. Also, Objective Structured Clinical Examinations (OSCEs) were used to grade all residents 2 weeks after their sessions. RESULTS: All pre-post changes were calculated and no statistically significant differences were seen (P > 0.50). OSCE interpersonal skills scores and content scores were calculated. The interactive group had statistically significantly higher scores (P < 0.05). CONCLUSION: The data supports the conclusion that an interactive method is more effective to use to train HIV risk assessment interview skills to medical residents when compared to the didactic method.

Error budgets: a system to characterize error source in health care experimentation.

OBJECTIVE: In the design of an experiment, it is important that all significant sources of error be identified and characterized accurately before the experiment is executed. For many experiments, the experiment designer is less inclined to ignore significant error factors when he adopts the use of an error budget in his experiment design process. A plan for generating an accurate error budget is presented along with a discussion concerning the disposition of significant error sources. METHODS: Specific methods discussed to generate the error budget are the fishbone diagram, factor weighing and Pareto analysis. Specific dispositions suggested are to eliminate, attenuate, characterize or to simply acknowledge the error with no remedy. These procedures, adhered to by most successful experiment design teams in industry, have evolved over recent decades and seem to be of use as they are applied to health care research. RESULTS: Use of the outlined method forces the experimenter to become familiarized with the most significant variables of the experiment he or she is designing through the deliberate examination of experimental noise contributors and through the comparison of the experimental noise to the magnitudes of the measurements in the experiment. CONCLUSION: The design of health care research experiments may be streamlined by incorporating the methods outlined in the generation of an error budget and by properly using the error budget to dispose of each source of error properly.