Tag Archives: Hygiene

Important New Data on WASH and Nutritional Interventions from Kenya and Bangladesh

Two recent cluster RCTs published in Lancet Global Health have reported on WASH (water quality, sanitation and handwashing) and nutritional interventions.[1] [2] One was conducted in Bangladesh and the other in Kenya; they were both based on recruitment of pregnant woman, who were then grouped in delineated clusters based on geographical proximity. The studies were all entirely rural.

Both studies found that nutritional interventions, either singly or in combination with other wash interventions, improved child growth. The Kenyan study found no benefit for WASH interventions on reported rates of diarrhoea. However, the Bangladesh study did find a reduction of diarrhoea of about 40% in the WASH intervention groups. A follow-on study to the Bangladesh trial at one year of age found an improvement in developmental milestones across all of the intervention groups.[3]

Probably the strongest message to come out here, is that nutritional interventions improve growth in deprived rural populations. This is consistent with most, but not all, previous literature. The discordant effects of WASH interventions on diarrhoea rates across Kenya and Bangladesh is puzzling and no convincing explanation is offered by the authors. I note, however, that the prevalence of diarrhoea was much higher in Kenya than in Bangladesh. Of course, that provides more headroom for improvement In the Kenyan setting, making the discordant results even more perplexing. One possibility, is that reported diarrhoea rates are just a very poor marker for gastrointestinal disease. Worse still, they are ‘reactive’, meaning that if people are aware that they are on the receiving end of an intervention to reduce diarrhoea rates, then they may report less diarrhoea, even if the true prevalence is unchanged.[4] We are investigating this possibility in a study in Mwanza, Tanzania, which we are conducting in collaboration with UN-Habitat. I am not sure how to interpret these results with respect to the theory that chronic gastrointestinal infections aggrevate malnutrition by causing a chronic malabsorptive small bowel enteropathy.

It is interesting to compare these results with the effects of the sanitary revolution in Europe and North America over a century ago.[5] Here again, water and sanitation had modest and inconsistent effects on childhood diarrhoea, but with much more dramatic effects on typhoid and cholera. Taken in the round these recent results reported in Lancet Global Health are consistent with historical data.

— Richard Lilford, CLAHRC WM Director

References:

  1. Luby SP, Rahman M, Arnold BF, et al. Effects of water quality, sanitation, handwashing, and nutritional interventions on diarrhoea and child growth in rural Bangladesh: a cluster randomised controlled trial. Lancet Glob Health. 2018; 6: e302-15.
  2. Null C, Stewart CP, Pickering AJ, et al. Effects of water quality, sanitation, handwashing, and nutritional interventions on diarrhoea and child growth in rural Kenya: a cluster-randomised controlled trial. Lancet Glob Health. 2018; 6: e316-29.
  3. Tofail F, Fernald LCH, Das KK, et al. Effects of water quality, sanitation, hand washing, and nutritional interventions on child development in rural Bangladesh (WASH Benefits Bangladesh): a cluster-randomised controlled trial. Lancet Child Adolesc Health. 2018; 2: 255-68.
  4. Clasen T, Boisson S, Routray P, Torondel B, Bell M, Cumming O, et al. Effectiveness of a rural sanitation programme on diarrhoea, soil-transmitted helminth infection, and child malnutrition in Odisha, India: a cluster-randomised trial. Lancet Glob Health. 2014; 2(11): e645-53.
  5. Szreter S. The Population Health Approach in Historical Perspective. Am J Public Health. 2003; 93(3): 421-31.
Advertisements

Evaluation of High vs. Low Cost Service Interventions

Generic service interventions vary considerably in their costs. Human resource interventions, such as improving the nurse to patient ratio or making more specialists available over the weekend, tend to be expensive. Other service interventions, such as an educational intervention to improve team working in multi-disciplinary clinical teams, are less expensive. The cost-effective effect size is smaller for lower cost interventions than for those which are more expensive. The axiom, that the cost of an intervention determines the effectiveness threshold at which it becomes cost-effective, has profound implications for the design and analysis of evaluative studies. The nub of the argument is that the size of the health effect that would justify deployment of service interventions may be too small to be detected by affordable or logistically feasible studies when the cost of that intervention is low. Before developing this argument further, let me be clear that by cost I mean net cost (not just the cost of the intervention itself), and that costs must be compared with respect to a common denominator – e.g. cost per patient, cost per 1,000 patients, etc.

Let us imagine that we wish to improve consultant cover at weekends. This is a very expensive intervention (whether measured in terms of the cost of hiring new consultants or the opportunity costs of re-allocating consultant time).[1] Such an intervention would need to provide considerable health gain to justify its substantial cost. In such a case it is reasonable to expect – indeed require – that any evaluative study should be able to detect patient benefit, say in terms of lives saved and adverse events avoided. If no improvement in health gain is detected, then we must conclude that either the study was ‘underpowered’ OR that any effects are too small to justify the intervention costs. If the study was not underpowered – that is to say if the sample size was sufficient to detect health benefits sufficient to justify cost of the intervention – then we conclude that the intervention does not promise good value for money. We leave aside the issue of exactly how the threshold effect (which justifies an intervention cost) can be determined, save to point out that methods to do so exist and that we have advocated use of such methods (prospective health economic modelling) for sometime.[1-3]

Take, as an opposite extreme, an intervention to promote hand-washing – perhaps using ‘nudge theory’. The intervention here is likely to be nugatory – say having a sticker with an illustration of a ‘watching eye’ placed over hospital sinks, for example.[4] Harms are unlikely and intervention costs are low. It follows that there is not much downside to intervening. That is to say, even if the intervention was totally ineffective, no real harm would result. A massive trial with an endpoint such as hospital-acquired infection rates would be overkill in such a scenario. This is because the threshold effect to justify the intervention is much smaller than the minimal difference detectable in any affordable / logistically feasible study. Using ‘upstream’ endpoints, such as “was the intervention deployed?” and “did it increase use of hand-washing materials?” (necessary but not sufficient conditions for effectiveness) would suffice in an evaluation. Many interventions are rather more expensive than promotion of hand-washing, but much less expensive than large HR initiatives; the above mentioned educational intervention to promote team-work, for example. Here it might be too much to expect, or require, quality of life or mortality to change sufficiently for any change to be detectable (statistically) in an affordable trial. However, one might expect to pick up a broad range of other signals that an intervention effect was likely. For example, it may be observed that team working and patient satisfaction had improved, as well as that the intervention was adopted and supported by staff. That one might have to rely on such proxies for QoL and life years has been referred to as “an inconvenient truth in service delivery research.”[5] It is important that grant awarding panels should not follow a one-size fits all approach to service delivery research, but rather that they should tailor their requirements according to the cost of intervention concerned. Likewise, they should be prepared to integrate many sources of evidence in their assessment of health benefit parameters, as argued elsewhere,[1] [6-8] and in the report of a recent study later in this issue of your News Blog.

— Richard Lilford, CLAHRC WM Director

References:

  1. Sutton M, Birbeck SG, Martin G, Meacock R, Morris S, Sculpher M, Street A, Watson SI, Lilford RJ. Economic analysis of service and delivery interventions in health care. Health Serv Del Res. 2018; 6(5).
  2. Girling A, Liflord R, Cole A, Young T. Headroom Approach to Device Development: Current and Future Directions. Int J Technol Assess Health Care. 2015; 31(5): 331-8.
  3. Yao GL, Novielli N, Manaseki-Holland S, Chen YF, van der Klink M, Barach P, Chilton PJ, Lilford RJ, European HANDOVER Research Collaborative. Evaluation of a predevelopment service delivery intervention: an application to improve clinical handovers. BMJ Qual Saf. 2012; 21(s1):i29-i38.
  4. King D, Vlaev I, Everett-Thomas R, Fitzpatrick M, Darzi A, Birnbach DJ. “Priming” Hand Hygiene Compliance in Clinical Environments. Health Psychol. 2016; 35(1): 96-101.
  5. Lilford RJ, Chilton PJ, Hemming K, Girling AJ, Taylor CA, Barach P. Evaluating policy and service interventions: framework to guide selection and interpretation of study end points. BMJ. 2010; 341: c4413.
  6. Watson SI & Lilford RJ. Essay 1: Integrating multiple sources of evidence: a Bayesian perspective. In: Challenges, solutions and future directions in the evaluation of service innovations in health care and public health. Southampton (UK): NIHR Journals Library, 2016.
  7. Watson SI, Chen YF, Bion JF, Aldridge CP, Girling A, Lilford RJ; HiSLAC Collaboration. Protocol for the health economic evaluation of increasing the weekend specialist to patient ratio in hospitals in England. BMJ Open. 2018; 8(2): e015561.
  8. Lilford RJ, Girling AJ, Sheikh A, Coleman JJ, Chilton PJ, Burn SL, Jenkinson DJ, Blake L, Hemming K. Protocol for evaluation of the cost-effectiveness of ePrescribing systems and candidate prototype for other related health information technologies. BMC Health Serv Res. 2014; 14:314

Three Hits Hypothesis

Quite a lot of diseases are brought about by the conflation of two factors. Mice infected with certain herpes viruses suffer no ill-effect unless a helminth infestation supervenes. Oral allergy syndrome arises when a certain pollen interacts with certain foods (usually raw fruits, vegetables and nuts). The hygiene hypothesis says that lack of exposure to certain gut bacteria sensitises the body to allergic reactions to a range of environmental allergens. The pathway for disease involves three hits:

Genetically predisposed person –> Exposure 1 –> Exposure 2 –> Disease.

An intriguing example of a three-hit condition is the severe disease of children – Burkitt’s lymphoma. This cancer arises in germinal centres of lymph nodes in the neck. It is known that Epstein-Barr (EB) virus infection is necessary for endemic Burkitt’s lymphoma to develop because it prevents apoptosis (cell death) when certain mutations occur in the cell. But endemic Burkitt’s lymphoma only occurs in the malaria belt, and why this is so has been a mystery until the last few years. Now we know that the malaria parasite Plasmodium falciparum ‘upregulates’ an enzyme that causes mutations in DNA in lymph cells. These mutations are a normal part of antibody production since rearrangements of chromosome segments is necessary for antibody specificity. But in people with falciparum malaria, the effect ‘spills over’ to cause mutations of cancer genes. The double hit of EB plus malaria sets the scene for carcinogenesis.[1] Why in the neck – perhaps because lymph cells in the necks of children work particularly hard eradicating throat and ear infections, in which case there is a ‘four hits’ hypothesis!

— Richard Lilford, CLAHRC WM Director

References:

  1. Thorley-Lawson D, Deitsch KW, Duca KA, Torgbor C. The Link between Plasmodium falciparum Malaria and Endemic Burkitt’s Lymphoma—New Insight into a 50-Year-Old Enigma. PLoS Pathog. 2016; 12(1): e1005331.

Fine Dining and Fine Hygiene are Negatively Correlated

A recent study shows that restaurants rated highly in food guides are associated with a greater overall risk of foodborne gastrointestinal diseases outbreaks than your run-of-the-mill restaurant.[1] However, the ‘high-end’ restaurants also score more highly on the Food Agency Inspection visits. So why do the posh restaurants generate more GI diseases than their more mundane peers despite better hygiene in the restaurants with the best food? The high disease risk in highly rated restaurants probably comes from the nature of the food served (e.g. oysters) and cooking methods (e.g. low temperatures to produce chicken liver parfait). So the risk is real, but worth running!

— Richard Lilford, CLAHRC WM Director

Reference:

  1. Kanagarajah S, Mook P, Crook P, Awofisayo-Okuyelu A, McCarthy N. Taste and Safety: Is the Exceptional Cuisine Offered by High End Restaurants Paralleled by High Standards of Food Safety? PLoS Curr Outbreaks. 2016.