Anti-tuberculosis medication side-effects constitute major factor for poor adherence to tuberculosis treatment

ABulletin of the World Health Organization | March 2008, 86 (3) Letters Reaching Every District (RED) approach: a way to improve immunization performance In their paper, Victora et al.1 show that “child survival interventions are inequitably distributed within low- and middle-income countries”. Areas of greatest need were not prioritized, and expansion of these health programmes in more difficult areas has tended to be delayed or postponed. In response, we wish to share some results and propose a way forward based upon experiences with immunization programmes. Immunization programmes around the world have recognized and strived to reduce inequity for many years. While Universal Child Immunization (UCI) of 80% coverage was achieved in 1990, this merely emphasized the need to balance the inequalities within and between countries. Accordingly, several approaches were adopted. The “high risk approach” was designed in the mid- 1990s to reach women in underserved areas with tetanus toxoid immuniza- tion using a campaign-style approach.2 District level microplanning has been the cornerstone of the polio eradica- tion and measles elimination initiatives, to maximize the delivery of vaccines to all districts, especially underserved populations. District-level coverage and disease surveillance data are now routinely collected in most countries, with reporting of selected indicators to the global level since 2000. In 2002, the Reaching Every District (RED) approach was developed and introduced by WHO, the United Nations Children’s Fund (UNICEF) and other partners in the GAVI Alliance to improve immunization systems in areas with low coverage. Far from being a programme, or separate initiative, the approach outlines five operational components that are specifically aimed at improving coverage in every district: re-establishment of regular outreach • services; a Health Section, Programme Division, UNICEF, New York, NY, United States of America. b Department of Immunization and Biologicals, World Health Organization, 20 avenue Appia, 1211 Geneva 27, Switzerland. c Regional Office for Africa (AFRO), World Health Organization, Brazzaville, Republic of the Congo. Correspondence to Jos Vandelaer (e-mail: vandelaerj@who.int). supportive supervision: on-site • training; community links with service • delivery; monitoring and use of data for • action; better planning and management • of human and financial resources.3 The RED approach encourages coun- tries to use coverage data to make an analysis of the distribution of unim- munized infants, and thereby prioritize districts with poor access and utiliza- tion of immunization, while districts are encouraged to make microplans to identify local problems and adopt cor- rective solutions. Since 2003, 53 developing countries have started implementing RED to various degrees, mostly in Africa and south and south-east Asia.4 All 53 countries belong to the groups of lower income and lower-middle income countries, as per World Bank classification. In 2005, an evaluation of 5 countries in Africa that had imple- mented RED found that, in 4 of the 5 countries, immunization coverage had increased since the implementation of RED, and that the proportion of districts with DTP3 (three-dose diph- theria, tetanus and pertussis vaccine) coverage above 80% had more than doubled.5 The number of unimmu- nized children in these 5 countries was reduced from 3 million in 2002 to 1.9 million in 2004. Interestingly, the re- port notes that outreach services, one of the five components of RED, were of- ten used to deliver other interventions beyond immunization, such as Vitamin A, antihelminthic drugs or insecticide- treated bed nets. This indicates that implementation of RED components may start to have an impact beyond immunization services alone. An analysis of coverage data sup- ports the findings of the evaluation in Africa. It shows that in the 53 countries that started to implement RED be- tween 2003 and 2005, DTP3 coverage (as estimated by WHO and UNICEF) increased between 2002 and 2005 in 34 (64%) countries, and decreased in only 7 (13%).6 Although these data need to be interpreted with caution, since RED implementation has not been nation- wide in many countries, they seem to indicate that where RED is imple- mented, it can help to reduce gaps in immunization coverage. We agree with the suggestion of Victora et al. regard- ing the need for information systems and training. Most of the 53 countries we refer to have functional immuniza- tion information, logistics and supply systems and have implemented district training, often using funds from the GAVI Alliance. Furthermore WHO, UNICEF and other partners at country and regional level have been closely involved in guiding countries adopt- ing the RED approach to reach the unreached. We believe that the RED approach of district microplanning based upon local data using simple op- erational components and supported by supply and logistics has the poten- tial for the successful delivery of other child health interventions, especially during outreach. ■ Jos Vandelaer,a Julian Bilousb & Deo Nshimirimanac References Victora CG, H1. uicho L, Amaral JJ, Armstrong- Schellenberg J, Manzi F, Mason E et al. Are health interventions implemented where they are most needed? District uptake of the Integrated Management of Childhood Illness strategy in Brazil, Peru and the United Republic of Tanzania. Bull World Health Organ 2006;84:792-801. PMID:17128359 doi:10.2471/BLT.06.030502 Anonymous. The high-risk approach: the WHO-2. recommended strategy to accelerate elimination of neonatal tetanus. Wkly Epidemiol Rec 1996;71:33-9. PMID:8851826 Global3. Immunization Vision and Strategy 2006-2015. Geneva: WHO and UNICEF; 2005. Global4. Polio Eradication Initiative: 2005 annual report. Geneva: WHO, Rotary International, CDC, UNICEF; 2006 (WHO/Polio/06.02). Letters B Bulletin of the World Health Organization | March 2008, 86 (3) Report5. of evaluation of Reaching Every District approach in five countries. Brazzaville: WHO Regional Office for Africa, 2005 [unpublished document]. WHO vaccine-preventable diseases: monitoring 6. system: 2006 global summary. Geneva: WHO; 2006 (WHO/IVB/2006). Available from: http:// www.who.int/immunization_monitoring/data/en/ Health insurance in sub- Saharan Africa: a call for subsidies De Allegri et al.1 rightly describe low enrolment as a principal problem re- lated to the functioning of community health insurance (CHI) in sub-Saharan Africa. Furthermore, they identify a set of important factors affecting the decision to enrol. Nonetheless, on re- flection about the evidence established by this paper and related research, I would like to suggest some additional considerations. First of all, the described (and not too surprising) fact that the education- al and, importantly, the socioeconomic status of a household play predomi- nant roles in the decision of whether to enrol in health insurance is depicted by a series of articles2 as well as several systematic article reviews.3 Some of them are quoted by the authors them- selves.4,5 Second, the consistency of this ob- servation and the clear-cut cause–effect relationship between socioeconomic well being and the readiness to embark on an expenditure (be it for health insurance or anything else) allow the conclusion that wealth is a fundamental predictive factor for enrolment into health insurance. Third, if then poverty can be under- stood as a risk factor for not embarking into health insurance, the discussion around an insurance approach for the poor should focus very much on the fol- lowing three questions: What percent- age of the population targeted by the envisaged or existing insurance scheme are too poor to enrol on their own? By which kind of corrective measures can they be included? What consequences do these measures have for the financial viability of the scheme? a Health Sector Coordinator, German Cooperation, GTZ, BP 59, Kigali, Rwanda. Correspondence to Andreas Kalk (e-mail: andreas.kalk@gtz.de). Two recent analyses from Ghana6 and Rwanda7 suggest that the capacity of households to contribute financially is so weak that the dual objectives of mobilizing significant resources for health on one side, and of covering a large percentage of the targeted rural population on the other, are mutually exclusive. That is to say that insur- ance schemes requiring a contribution of little more than a few US dollars per year are beyond the reach of the majority, but they still do not allow the financing of reasonable (and thus attractive) health services! Furthermore, schemes charging about ten times such an amount are still affordable by a con- siderable minority of the population and maximize resource mobilization in absolute terms. This phenomenon is explained largely by the highly skewed distribution of wealth in the settings studied (as expressed equally by a high Gini coefficient). This finding seems to be one of the main reasons underlying the aforementioned low enrolment rate scrutinized by De Allegri et al. In many countries in sub-Saharan Africa, health insurance schemes might find them- selves in a tragic situation: Depending on the design, people are either unable to pay for the schemes, or the schemes are unable to pay for the envisaged services. Therefore, it is suggested that future research go beyond the identifi- cation of additional predictive factors for health insurance enrolment. If health insurance is to cover broader population strata in sub-Saharan Africa and to assure satisfactory health services, schemes will require continu- ous and long-term subsidies to bridge the gap between household capacity to contribute financially and the real costs of health care. The development of approaches addressing this dilemma should be considered as a research priority. They might include initiatives of north–south risk pooling as between the Netherlands and Ghana.8 This ne- cessity is underpinned by the capacity of health insurance to formalize social protection and to create a market between health service providers and their “customers”, simultaneously alleviating poverty and empowering communities. Yet, available evidence points out that to play these roles, health insurance needs subsidies. ■ Andreas Kalk a References De Allegri M, Kouyaté B, Becher H, Gbangou 1. A, Pokhrel S, Sanon M. et al. Understanding enrolment in community health insurance in sub-Saharan Africa: a population-based case- control study in rural Burkina Faso. Bull World Health Organ 2006;11:852-8. Musango L, Dujardin B, Dramaix M, Criel B. 2. Les profils des membres et non membres des mutuelles de santé au Rwanda: le cas du district sanitaire de Kabutare. Trop Med Int Health 2004;9:1222-7. PMID:15548320 doi:10.1111/ j.1365-3156.2004.01318.x Preker AS, Carrin G, Dror D, Jakab M, Hsiao W, 3. Arhin-Tenkorang D. Effectiveness of community health financing in meeting the cost of illness. Bull World Health Organ 2002;80:143-50. PMID:11953793 Ekman B. Community-based health insurance in 4. low-income countries: a systematic review of the evidence. Health Policy Plan 2004;19:249-71. PMID:15310661 doi:10.1093/heapol/czh031 Walkens MP, Criel B. 5. Les mutuelles de santé en Afrique sub-Saharienne – Etat de lieu et réflexion sur un agenda de recherche. Washington, DC: World Bank; 2004 [Health, Nutrition and Population Discussion Paper]. Arhin-Tenkorang D. Experience of community 6. health financing in the African region. In: Health financing for poor people: resource mobilization and risk sharing. Washington, DC: World Bank; 2004. Schmidt JO, Mayindo JK, Kalk A. Thresholds for 7. health insurance in Rwanda: who should pay how much? Trop Med Int Health 2006;11:1327-33. PMID:16903895 doi:10.1111/j.1365- 3156.2006.01661.x Improving social protection for the poor: 8. health insurance in Ghana – The Ghana social trust pre-pilot project. Geneva: International Labour Organization; 2005. Anti-tuberculosis medication side-effects constitute major factor for poor adherence to tuberculosis treatment Two significant issues that require further clarification in Garner et al.’s stimulating paper (Promoting adherence to tuberculosis treatment 1) are the impact of medication side-effects on treatment adherence as well as how adherence to tuberculosis (TB) chemotherapy should be defined and monitored. The treatment regimen recommended Letters CBulletin of the World Health Organization | March 2008, 86 (3) within the DOTS approach is as- sociated with significant side-effects. Side-effects such as hepatitis, dyspepsia, exanthema and arthralgia were respon- sible for termination of therapy in up to 23% of patients during the intensive phase.2 Medication side-effects were also found to be significantly associ- ated with defaulting.3 At Kyrgyzstan prisons, where the author worked as a TB doctor in early 2007, medication side-effects were among the most com- mon reasons for patient non-attendance at DOTS clinics. The author observed similar non-attendance and defaulting trends among community-based TB patients in northern Nigeria during the 1990s. The side-effects profile of TB chemotherapy is magnified in patients with concurrent HIV treatment and/ or prior history of hepatitis,4 and those being treated with second-line drugs for multidrug-resistant TB, during which as many as 86% of patents may develop medication side-effects.5,6 To minimize the adverse impact of medication side- effects in TB treatment adherence, it is important that TB health staff are adequately trained on their recogni- tion and management. Such training should include how to provide concise pretreatment counselling to patients on possible side-effects of treatment.7 It is also important that medications for managing side-effects should be ordered concurrently with the ordering of anti-TB chemotherapy to facilitate timely and adequate treatment of such side-effects. The DOTS strategy contains ele- ments of adherence and compliance. While these terms were initially used synonymously and are still commonly used interchangeably in TB literature, they have subtle but noteworthy signifi- cant differences. The term “adherence” (or “patient-centred compliance”8) refers to the extent to which patients follow a prescribed regimen. It im- plies a more active and collaborative involvement of patients working with health-care providers in managing their treatment. “Adherence” is currently preferred to “compliance” in medical literature as it portrays a more respectful and active role of the patient in disease management. It captures the increas- ing complexity of TB chemotherapy by characterizing patients as independent, intelligent and autonomous people who take active and voluntary roles in defin- ing and pursuing goals for their medical treatment. The extent of treatment adherence may be facilitated by positive or negative attributes related to health system, social/family issues, personal fac- tors, and drug factors (e.g. medication side-effects are negative drug attributes while a fixed-dose combination is a positive drug attribute in relation to treatment adherence). Empowerment of people with TB, and communities, through advocacy, communication and social mobilization as well as patient and community participation in TB care are important in facilitating treatment adherence using the DOTS approach.9 In exceptional situations, the DOTS approach of facilitating adher- ence might not achieve its objectives, since patients need to make themselves available for treatment and are less likely to do so if they are imprisoned, suffer medication side-effects or experi- ence homelessness, drug addiction, unemployment or alcoholism.4,10 In Kyrgyzstan prisons, the practice of self- administered anti-TB treatment on weekends was discontinued in March 2007 due to repeated documented evidence that many patients were trafficking their weekend TB medica- tions, despite concerted efforts aimed at enhancing patient empowerment and peer support. The most cited definition of treat- ment compliance is by Haynes – “the extent to which a person’s behaviour (in terms of taking medication following diets, or executing lifestyle changes) co- incides with medical or health advice”.11 “Compliance” may be used to describe the “right of public health authorities to demand adherence”1 such as by compel- ling patients to take TB chemotherapy using Public Health Detention Orders.12 Or it may be used as a framework to evaluate adherence. For example, patients who adhere to TB medication as prescribed 95% of the time are said to demonstrate high compliance, while patients who adhere for 40% of the time are said to demonstrate low compliance. The World Health Organization defines a TB treatment defaulter as a patient whose treatment was inter- rupted for two consecutive months or more. It indicates a closure of the current treatment, and documents that patients’ compliance has been 0% for so long. As with HIV treatment, TB therapy requires high (> 90%) compli- ance to facilitate cure. Good adherence results in high compliance and absence of treatment default. Default rate is a crude approach to adherence monitor- ing, since it does not really reveal why the patient interrupted treatment for 2 or more consecutive months. Promptly implementing compliance assurance measures provide for better adherence monitoring than defaulter tracing, pro- vided that a baseline compliance level is set at which investigation of the reasons for poor adherence can be undertaken. Currently at the Kyrgyzstan prison TB project, we undertake investigation of reasons for poor adherence if a patient misses at least two doses of anti-TB treatment in a week. This “patient con- tact” baseline period is in line with the maximum duration of non-adherence that will adversely impact on the ef- ficacy of treatment.13 ■ Niyi Awofesoa References Garner P, Smith H, Munro S, Volmink J. 1. Promoting adherence to tuberculosis treatment. Bull World Health Organ 2007;85:404-6. PMID:17639229 doi:10.2471/BLT.06.035568 Schaberg T, Rebham K, Lode H. Risk factors 2. for side-effects of isoniazid, rifampicin and pyrazinamide in patients hospitalized for pulmonary tuberculosis. Eur Respir J 1996;9:2026-30. PMID:8902462 doi:10.1183/ 09031936.96.09102026 Tekle B, Mariam DH, Ali A. Defaulting from 3. DOTS and its determinants in three districts of Arsi zone in Ethiopia. Int J Tuberc Lung Dis 2002;6:573-9. PMID:12102295 Fry RS, Khoshnood K, Vdovichenko E, 4. Granskaya J, Sazhin V, Shpakovskaya L, et al. Barriers to completion of tuberculosis treatment among prisoners in St. Petersburg, Russia. Int J Tuberc Lung Dis 2005;9:1027-33. PMID:16158896 Torun T, Gungor O, Ozmen I, Bolukbasi Y, 5. Maden E, Bicakci B, et al. Side effects associated with the treatment of multi-drug resistant tuberculosis. Int J Tuberc Lung Dis 2005;9:1373-7. PMID:16468160 a School of Public Health & Community Medicine, University of New South Wales, Sydney, NSW, Australia. Correspondence to Niyi Awofeso (e-mail: niyi.awofeso@justicehealth.nsw.gov.au). Letters D Bulletin of the World Health Organization | March 2008, 86 (3) Leimane V, Riekstina V, Holtz TH, Zarovska V, 6. Skripconoka L, Thorpe K, et al. Clinical outcome of individualized treatment of multi-drug resistant tuberculosis in Latvia: a retrospective cohort study. Lancet 2005;365:318-26. PMID:15664227 Self-study modules on tuberculosis: patient 7. adherence to tuberculosis treatment. Washington, DC: National Center for HIV, STD and TB Prevention, US Department of Health and Human Services; 1999. Available from: http://www.cdc.gov/tb/pubs/ssmodules/pdfs/9. pdf Conrad P. The meaning of medication: another 8. look at compliance. Soc Sci Med 1985;20:29-37. PMID:3975668 doi:10.1016/0277- 9536(85)90308-9 Building on and enhancing DOTS to meet the 9. TB-related Millennium Development Goals. Geneva: WHO: 2006. Available from: http:// www.who.int/tb/publications/2006/who_htm_ tb_2006_368.pdf Khan MA, Walley JD, Witter SN, Shah SK, 10. Javeed S. Tuberculosis patient adherence to direct observation: results of a social study in Pakistan. Health Policy Plan 2005;20:354-65. PMID:16183735 doi:10.1093/heapol/czi047 Haynes RB. Determinants of compliance: the 11. disease and the mechanics of treatment. In: Compliance in Health. Baltimore: Johns Hopkins University Press; 1979. Senanayake SN, Ferson MJ. Detention for 12. tuberculosis: public health and the law. Med J Aus. 2004;180:573-6. Frieden T. What is intermittent treatment and 13. what is the scientific basis for intermittency. In: Toman’s tuberculosis: case detection, treatment and monitoring. Geneva: WHO; 2004. Response to opt-out approach to prevent mother-to-child transmission of HIV I read with great interest the paper on routine HIV testing for pregnant women in Zimbabwe by Winfreda Chandisarewa et al.1 The paper reported a significant increase in the acceptance of PMTCT (preventing mother-to- child transmission) services such as HIV testing, counselling and follow-up after the introduction of routine HIV testing (“opt-out” approach). However, the conclusion that the opt-out approach for HIV testing was operationally feasible and acceptable to all women, and that HIV-infected women reported relatively low levels of spousal abuse and other adverse social consequences, seems to be overstated. a Bureau of International Cooperation, International Medical Center of Japan, Toyama 1-21-2, Shinjuku-city, Tokyo, Japan. Correspondence to Kazuhiro Kakimoto (e-mail: k-kakimoto@it.imcj.go.jp). Results from a survey of the tested mothers indicated that approximately 10% of those women who disclosed their test results still experienced negative effects. These findings hardly indicate “low levels of spousal abuse and other adverse social consequences” in consideration of these women’s personal safety. It is necessary to evaluate the opt- out approach for HIV testing by com- paring the incidence of adverse effects with mothers who opted-in to HIV testing to conclude whether the benefit of the opt-out approach outweighs the risk. More importantly, more attention must be paid to the issue of domestic violence by partners after disclosure, since many cases have been reported in African countries.2–4 In addition, the authors should show the percentage of those mothers who had been tested and counselled before the study period. As the majority of mothers in the study were reportedly multiparae, they might have been tested for HIV in previous pregnancies. Mothers re-tested during the study are likely to have experienced fewer negative effects. Moreover, I would like to suggest that the authors provide more infor- mation on the role of the community mobilization activities conducted before the introduction of the opt-out approach. Barriers and predictors to HIV testing have been investigated to improve the acceptance of HIV testing in PMTCT services.5,6 This research shows that community activities play an important role in clearing some of the barriers to testing and counselling ser- vices, therefore providing an entry point to prevention and care services includ- ing PMTCT. These activities, together with high-quality counsellors, might have contributed to increasing the ac- ceptance of HIV testing and counselling and to reducing its adverse effects. The contents of the community mobilization activities, including male involvement, could be analysed more and shared with readers, so as to provide a good model to commencing provider- initiated HIV testing and counselling (PITC) in other areas. There have been a lot of argu- ments about human rights and HIV testing. However, I would like to stress that we need more good practices with successful increased uptake of PMTCT services and minimal negative impact, so as to provide practical ideas for the adaptation of the WHO guidelines on PITC at country level.7 The activities outlined in this study, especially those conducted in the community, could help to provide such ideas. ■ Kazuhiro Kakimotoa References Chandisarewa W, Stranix-Chibanda L, Chirapa E, 1. Miller A, Simoyi M, Mahomva A, et al. Routine offer of antenatal HIV testing (“opt-out” approach) to prevent mother-to-child transmission of HIV in urban Zimbabwe. Bull World Health Organ 2007;85:843-50. PMID:18038074 Baiden F, Remes P, Baiden R, Williams J, 2. Hodgson A, Boelaert M, et al. Voluntary counseling and HIV testing for pregnant women in the Kassena-Nankana district of northern Ghana: is couple counseling the way forward? AIDS Care 2005;17:648-57. PMID:16036251 Medley A, Garcia-Moreno C, McGill S, Maman S. 3. Rates, barriers and outcomes of HIV serostatus disclosure among women in developing countries: implications for prevention of mother- to-child transmission program. Bull World Health Organ 2004;82:299-307. PMID:15259260 Maman S, Mbwambo JK, Hogan NM, Weiss E, 4. Kilonzo GP, Sweat MD. High rates and positive outcomes of HIV-serostatus disclosure to sexual partners: reasons for cautious optimism from a voluntary counseling and testing clinic in Dar es Salaam, Tanzania. AIDS Behav 2003;7:373-82. PMID:14707534 doi:10.1023/ B:AIBE.0000004729.89102.d4 Bajunirwe F, Muzoora M. Barriers to the 5. implementation of programs for the prevention of mother-to-child transmission of HIV: A cross- sectional survey in rural and urban Uganda. AIDS Res Ther 2005;2:10. doi:10.1186/1742- 6405-2-10 Maman S, Mbwambo J, Hogan NM, Kilonzo GP, 6. Sweat M. Women’s barriers to HIV-1 testing and disclosure: challenges for HIV-1 voluntary counselling and testing. AIDS Care 2001; 13:595-603. PMID:11571006 doi:10.1080/09540120120063223 Guidance on provider-initiated HIV testing and 7. counselling in health facilities. Geneva: WHO/ UNAIDS; 2007. Available from: http://whqlibdoc. who.int/publications/2007/9789241595568_ eng.pdf Letters EBulletin of the World Health Organization | March 2008, 86 (3) Fig. 1. Percent change in use of oral rehydration therapy during the two most recent Demographic and Health Surveys, 1992–2005 Burkina Faso, 1999–2003 Mali, 1996–2001 Colombia, 2000–2005 Eritrea, 1995–2002 Mozambique, 1997–2003 Malawi, 2000–2005 Bangladesh, 2000–2004 Peru, 1996–2000 Co un tr ie s Change in ORT use for children < 3 years of age (%) 0 –50 –40 –30 –20 –10 0 10 20 30 40 Haiti, 2000–2005 Viet Nam, 1997–2002 Egypt, 2000–2005 Rwanda, 2000–2005 United Republic of Tanzania, 1999–2004 Morocco, 1992–2003 Namibia, 1992–2000 Ghana, 1998–2003 Nepal, 1996–2001 Nicaragua, 1998–2001 Benin, 1996–2001 Dominican Republic, 1999–2002 Ethiopia, 2000–2005 Madagascar, 1997–2003 Zambia, 1996–2001 Jordan, 1997–2002 Bolivia, 1998–2003 Philippines, 1998–2003 Senegal, 1997–2005 Guinea, 1999–2005 Cameroon, 1998–2004 Uganda, 1995–2000 Indonesia, 1997–2002 Chad, 1997–2004 Nigeria, 1999–2003 Kenya, 1998–2003 Declines in case management of diarrhoea among children less than five years old In the January 2007 Bulletin, Forsberg et al. present an elegant analysis of the trends in diarrhoea case management using data from national Demographic and Health Surveys (DHS) from 1986 to 2003.1 They conclude that despite global efforts to promote appropriate diarrhoea management during this time, only slow progress has been made in the proportion of children treated with oral rehydration salts (ORS), recommended home fluids (RHF) or increased fluids. The proportion of children given continued feeding during diarrhoea episodes actually decreased. Forsberg et al. examined trends in diarrhoea management over all DHS surveys conducted from 1986 to 2003. Since the time horizon of their analysis encompasses the late 1980s and early 1990s, when national and global Control of Diarrhoeal Disease programmes were quite active, and the period in the second half of the 1990s when these transitioned into Letters F Bulletin of the World Health Organization | March 2008, 86 (3) a University at Buffalo, Buffalo, NY, United States of America. b US Agency for International Development, Washington, DC, USA. c Centers for Disease Control and Prevention, Atlanta, GA, USA, and Nairobi, Kenya. d Ministry of Health, Nairobi, Kenya. Correspondence to Pavani Kalluri Ram (e-mail: pkram@buffalo.edu). the Integrated Management of Child- hood Illness programme, we believe that the small increases in ORS, RHF and increased fluids reported do not accurately represent the current trend in diarrhoea management. To inform fu- ture programmatic strategies, we sought to determine the most recent trends in diarrhoea treatment by reviewing DHS data on management of diarrhoea among children < 3 years old in the 34 countries that conducted a DHS survey between 2000 and 2005 (avail- able at: http://www.measuredhs.com). We calculated the absolute difference between the percentage of children with diarrhoea who were reportedly given ORS, RHF or increased fluids during the most recent DHS survey (2000 to 2005) and the preceding DHS survey (1992 to 2000) and found declines in the use of ORS, RHF or increased flu- ids in 23 (68%) of 34 countries, ranging from < 1% (Rwanda) to 32% (Kenya and Nigeria) (Fig. 1). Eleven countries experienced declines greater than 10%. Children with diarrhoea who are not given ORS, RHF or increased flu- ids may be given the same amount of fluids as when they are well, a reduced amount of fluids or no fluids at all. In surveys conducted in or after 2000 in 43 countries, a median of 29% (range: 5–83%) of children actually received reduced or no fluids during a diarrhoea episode. Forsberg et al. report an annual decrease of 0.64% for this indicator. However, of 32 countries where this indicator was measured during both of the most recent surveys, 27 (91%) ex- perienced increases in the proportion of children < 3 years old receiving reduced or no fluids during diarrhoea episodes, with increases ranging from < 1 to 64% (median 10%). The median annual change in the proportion of children with diarrhoea receiving reduced or no fluids was an increase of 1.4% (range: –2% – 13%). Declines in use of rehydration seem to occur despite overall improve- ments in awareness of ORS. In the 40 countries having DHS surveys in or after 2000 and in which this indicator was measured, a median of 89% (range: 46–98%) of mothers of children with diarrhoea knew about ORS. Of 30 countries where this indicator was mea- sured during both of the most recent surveys, 17 (57%) experienced increases in the proportion of mothers aware of ORS. The median increase in the proportion of mothers aware of ORS was 11% (range: < 1–33%). Our analysis confirms Forsberg’s finding of a downward trend in the practice of continued feeding during diarrhoea. In the 42 countries having DHS surveys conducted in or after 2000 and in which this indicator was measured, a median of 52% (range: 37–67%) of children received reduced or no food during a diarrhoea episode. Of 30 countries where this indicator was measured during both of the most recent surveys, 17 (57%) experienced increases in the proportion of children < 3 years old receiving reduced or no food during diarrhoea episodes. The median increase in the proportion of children with diarrhoea getting reduced or no food was 8% (range: 2–31%). The decline in appropriate diar- rhoea case management at the house- hold level is likely multifactorial. The well-deserved recent growth in attention and resources accorded to diseases such as HIV/AIDS, tuberculosis and malaria has not been matched for other leading causes of childhood death, including diarrhoea.2 The shift away from verti- cal, disease-specific public health pro- grammes during the past decade towards more integrated approaches, which have been primarily implemented at health facilities and among health-care work- ers, may have resulted in gaps in promo- tion of basic diarrhoea case management at the community level.3 As Forsberg et al. point out, increased knowledge does not necessarily result in an improvement in practices. Thus, efforts to increase ap- propriate diarrhoea management must concentrate on behaviour change in the community and household, targeting a variety of caretakers involved in treat- ment decisions. To better understand these changes, we are undertaking quantitative and qualitative research to investigate the determinants of diarrhoea treatment by caregivers and health-care workers in Kenya, which has seen a substantial reduction in use of rehydration therapy. We encourage colleagues to undertake similar investigations in other countries showing evidence of declines in appro- priate diarrhoea management. We congratulate Forsberg et al. for highlighting the lack of progress in diarrhoea case management. Our additional findings underscore the dis- turbing fact that diarrhoea management behaviours are actually worsening in some countries. These findings indicate the possibility of losing ground on oral rehydration therapy, one of the simplest and most affordable public health in- novations of the past century. Without swift corrective action on multiple levels (community-based behaviour change, national and global funding and policy), we may indeed see reversals in the dramatic decline in diarrhoea mortality of the past 20 years, a decline frequently attributed to the advent of oral rehydration therapy.4 Renewed commitment to decreasing the highly preventable and treatable infant and child mortality from diarrhoea, which remains at 2 million deaths annually, is long overdue. ■ Pavani Kalluri Ram,a Misun Choi,b Lau- ren S Blum,c Annah W Wamae,d Eric D Mintzc & Alfred V Bartlett b References Forsberg BC, Petzold MG, Tomson G, Allebeck P. 1. Diarrhoea case management in low- and middle-income countries — an unfinished agenda. Bull World Health Organ 2007;85:42-8. PMID:17242757 Rudan I, El Arifeen S, Black RE, Campbell H. 2. Childhood pneumonia and diarrhoea: setting our priorities right. Lancet Infect Dis 2007;7:56-61. PMID:17182344 doi:10.1016/S1473- 3099(06)70687-9 Bryce J, Victora CG, Habicht JP, Black RE, 3. Scherpbier RW. Programmatic pathways to child survival: results of a multi-country evaluation of Integrated Management of Childhood Illness. Health Policy Plan 2005;20:i5-17. PMID:16306070 doi:10.1093/heapol/czi055 Victora CG, Bryce J, Fontaine O, Monasch R. 4. Reducing deaths from diarrhoea through oral rehydration therapy. Bull World Health Organ 2000;78:1246-55. PMID:11100619