Dengue disease surveillance: an updated systematic literature review.

OBJECTIVES: To review the evidence for the application of tools for dengue outbreak prediction/detection and trend monitoring in passive and active disease surveillance systems in order to develop recommendations for endemic countries and identify important research needs. METHODS: This systematic literature review followed the protocol of a review from 2008, extending the systematic search from January 2007 to February 2013 on PubMed, EMBASE, CDSR, WHOLIS and Lilacs. Data reporting followed the PRISMA statement. The eligibility criteria comprised (i) population at risk of dengue, (ii) dengue disease surveillance, (iii) outcome of surveillance described and (iv) empirical data evaluated. The analysis classified studies based on the purpose of the surveillance programme. The main limitation of the review was expected publication bias. RESULTS: A total of 1116 papers were identified of which 36 articles were included in the review. Four cohort-based prospective studies calculated expansion factors demonstrating remarkable levels of underreporting in the surveillance systems. Several studies demonstrated that enhancement methods such as laboratory support, sentinel-based reporting and staff motivation contributed to improvements in dengue reporting. Additional improvements for passive surveillance systems are possible by incorporating simple data forms/entry/electronic-based reporting; defining clear system objectives; performing data analysis at the lowest possible level (e.g. district); seeking regular data feedback. Six studies showed that serotype changes were positively correlated with the number of reported cases or with dengue incidence, with lag times of up to 6 months. Three studies found that data on internet searches and event-based surveillance correlated well with the epidemic curve derived from surveillance data. CONCLUSIONS: Passive surveillance providing the baseline for outbreak alert should be strengthened and appropriate threshold levels for outbreak alerts investigated. Additional enhancement tools such as syndromic surveillance, laboratory support and motivation strategies can be added. Appropriate alert signals need to be identified and integrated into a risk assessment tool. Shifts in dengue serotypes/genotype or electronic event-based surveillance have also considerable potential as indicator in dengue surveillance. Further research on evidence-based response strategies and cost-effectiveness is needed.

Human resource assessment for scaling up VL active case detection in Bangladesh, India and Nepal.

OBJECTIVES: To determine whether medical staff at PHC level would have the time to take up additional activities such as 1-day fever camps for active VL case detection. METHODS: This article assessed the workload of health staff of different professional categories working at health facilities in Bangladesh, India and Nepal. Data were collected from different sites in high endemic VL areas. The study population was the health staff of government health facilities at all levels. Workload indicators of staffing need (WISN) software were adopted to carry out the analysis of staff workload and their availability in the selected health facility. The WISN difference and WISN ratio for a particular health facility were calculated from actual staffing available and calculated staffing requirement. RESULTS: The results showed a mixed picture of the availability of health workers. In most settings of Bangladesh and India, physicians with or without laboratory technicians would have time for active case detection. In Nepal, this would be performed by trained nurses and paramedical personnel. CONCLUSION: If all vacant posts were filled, active case detection could be performed more easily. The elimination programme can be scaled up with the current staffing levels in the endemic areas with some short training if and when necessary.

Bacillus thuringiensis israelensis (Bti) for the control of dengue vectors: systematic literature review.

OBJECTIVE: To systematically review the literature on the effectiveness of Bacillus thuringiensis israelensis (Bti), when used as a single agent in the field, for the control of dengue vectors. METHOD: Systematic literature search of the published and grey literature was carried out using the following databases: MEDLINE, EMBASE, Global Health, Web of Science, the Cochrane Library, WHOLIS, ELDIS, the New York Academy of Medicine Gray Literature Report, Africa-Wide and Google. All results were screened for duplicates and assessed for eligibility. Relevant data were extracted, and a quality assessment was conducted using the CONSORT 2010 checklist. RESULTS: Fourteen studies satisfied the eligibility criteria, incorporating a wide range of interventions and outcome measures. Six studies were classified as effectiveness studies, and the remaining eight examined the efficacy of Bti in more controlled settings. Twelve (all eight efficacy studies and 4 of 6 effectiveness studies) reported reductions in entomological indices with an average duration of control of 2-4 weeks. The two effectiveness studies that did not report significant entomological reductions were both cluster-randomised study designs that utilised basic interventions such as environmental management or general education on environment control practices in their respective control groups. Only one study described a reduction in entomological indices together with epidemiological data, reporting one dengue case in the treated area compared to 15 dengue cases in the untreated area during the observed study period. CONCLUSION: While Bti can be effective in reducing the number of immature Aedes in treated containers in the short term, there is very limited evidence that dengue morbidity can be reduced through the use of Bti alone. There is currently insufficient evidence to recommend the use of Bti as a single agent for the long-term control of dengue vectors and prevention of dengue fever. Further studies examining the role of Bti in combination with other strategies to control dengue vectors are warranted.

Reviewing the development, evidence base, and application of the revised dengue case classification.

With the example of dengue, an evidence-based approach to prospectively develop a case classification is described, gathering evidence for identifying strength and weaknesses of the existing model, collecting new data describing the disease as it occurs globally, further developing a new model that can be applied in practice and field testing the newly developed model in comparison to the previous model. For each step in this process, the highest available level of evidence has been applied. This process has been initiated by the World Health Organization's (WHO) Special Programme for Research and Training in Tropical Diseases (TDR) and WHO's Department for Control of Neglected Tropical Diseases (NTD), developing the following for dengue. Since the early 1970s, dengue has been classified into dengue fever, dengue haemorrhagic fever grades I and II and dengue shock syndrome grades III and IV (DF/DHF/DSS). However, in recent years, a growing number of dengue clinicians have questioned the shortcomings of this scheme. The issues have revolved around the complexity of confirming DHF in clinical practice, misclassifying severe cases as DF, and the emphasis on haemorrhage rather than plasma leakage as the underlying problem in most severe dengue cases. Step 1: A systematic literature review highlighted the shortcomings of the DF/DHF/DSS scheme: (1) difficulties in applying the criteria for DHF/DSS; (2) the tourniquet test has a low sensitivity for distinguishing between DHF and DF; and (3) most DHF criteria had a large variability in frequency of occurrence. Step 2: An analysis of regional and national dengue guidelines and their application in the clinical practice showed a need to re-evaluate and standardize guidelines as the actual ones showed a large variation of definitions, an inconsistent application by medical staff, and a lack of diagnostic facilities necessary for the DHF diagnosis in frontline services. Step 3: A prospective cohort study in seven countries, confirmed the difficulties in applying the DF/DHF/DSS criteria even in tertiary care hospitals, that DF/DHF/DSS do not represent levels of disease severity and that a clear distinction between severe dengue (defined by plasma leakage and/or severe haemorrhage, and/or organ failure) and (non-severe) dengue can be made using highly sensitive and specific criteria. In contrast, the sub-grouping of (non-severe) dengue into two further severity levels was only possible with criteria that gave approximately 70% sensitivity and specificity. Step 4: Three regional expert consensus groups in the Americas and Asia concluded that 'dengue is one disease entity with different clinical presentations and often with unpredictable clinical evolution and outcome' and that, revising the results of Step 3, DF/DHF/DSS is not related to disease severity. Step 5: In a global expert consensus meeting at WHO in Geneva/Switzerland the evidence collected in Steps 1-4 was reviewed and a revised scheme was developed and accepted, distinguishing: dengue with or without warning signs and severe dengue; the further field testing and acquisition of further prospective evidence of the revised scheme was recommended. Step 6: In 18 countries, the usefulness and applicability of the revised classification compared to the DF/DHF/DSS scheme were tested showing clear results in favour of the revised classification. Step 7: Studies are under way on the predictive value of warning signs for severe dengue and on criteria for the clinical diagnosis of dengue which will complete the evidence foundation of the revised classification. The analysis has shown that the revised dengue case classification is better able to standardize clinical management, raise awareness about unnecessary interventions, match patient categories with specific treatment instructions, and make the key messages of patient management understandable for all health care staff dealing with dengue patients. Furthermore, the evidence-based approach to develop prospectively the dengue case classification could be a model approach for other disease classifications.

The Indian and Nepalese programmes of indoor residual spraying for the elimination of visceral leishmaniasis: performance and effectiveness.

Although, when applied under controlled conditions in India and Nepal, indoor residual spraying (IRS) has been found to reduce sandfly densities significantly, it is not known if IRS will be as effective when applied generally in these countries, via the national programmes for the elimination of visceral leishmaniasis. The potential benefits and limitations of national IRS programmes for the control of sandflies were therefore evaluated in the districts of Vaishali (in the Indian state of Bihar), Sarlahi (in Nepal) and Sunsari (also in Nepal). The use of technical guidelines, levels of knowledge and skills related to spraying operations, insecticide bio-availability on the sprayed surfaces, concentrations of the insecticide on the walls of sprayed houses, insecticide resistance, and the effectiveness of spraying, in terms of reducing sandfly densities within sprayed houses (compared with those found in unsprayed sentinel houses or control villages) were all explored. It was observed that IRS programme managers, at district and subdistrict levels in India and Nepal, used the relevant technical guidelines and were familiar with the procedures for IRS operation. The performance of the spraying activities, however, showed important deficiencies. The results of bio-assays and the chemical analysis of samples from sprayed walls indicated substandard spraying and suboptimal concentrations of insecticide on sprayed surfaces. This was particularly obvious at one of the Nepali study sites (Sunsari district), where no significant vector reduction was achieved. Sandfly resistance to the insecticide used in India (DDT) was widespread but the potential vectors in Nepal remained very susceptible towards a pyrethroid similar to the one used there. The overall short-term effectiveness of IRS was found to be satisfactory in two of the three study sites (in terms of reduction in the densities of the sandfly vectors). Unfortunately, the medium-term evaluation, conducted 5 months after spraying, was probably made invalid by flooding or lime plastering in the study areas. Preparation for, and the monitoring of, the IRS operations against sandfly populations in India and Nepal need to be improved.

Comparative study of kala-azar vector control measures in eastern Nepal.

This study was conducted to explore the most effective vector control tool among indoor residual spraying (IRS), long lasting insecticidal nets (LLINs) and ecological vector management (EVM) as a part of the regional visceral leishmaniasis elimination initiative. Alpha-cypermethrin as IRS, PermaNet as LLINs and plastering the inner walls of houses with lime as EVM were the interventions. One baseline and three follow-up entomological surveys were carried out in all arms using CDC miniature light traps (LT) and mouth aspirators. Comparisons were made between intervention arms and control arms with pre-intervention and post-intervention vector densities. Light traps were found more efficient in the collection of Phlebotomus argentipes in comparison with aspiration. Vector densities were significantly low in both IRS arm (p=0.009 in LT and p<0.001 in aspirator collections) and LLIN arm (p=0.019 in LT and p=0.023 in aspirator collections) in comparison with control arm. However, in EVM arm, there was no significant difference in P. argentipes sand fly density in comparison with control arm (p=0.785) in LT collections in follow-up surveys. Hence, IRS was found most effective control measure to decrease vector density. LLINs were also found effective and can be considered as a promising alternative vector control tool in VL elimination initiative.

Reducing costs and operational constraints of dengue vector control by targeting productive breeding places: a multi-country non-inferiority cluster randomized trial.

OBJECTIVES: To test the non-inferiority hypothesis that a vector control approach targeting only the most productive water container types gives the same or greater reduction of the vector population as a non-targeted approach in different ecological settings and to analyse whether the targeted intervention is less costly. METHODS: Cluster randomized trial in eight study sites (Venezuela, Mexico, Peru, Kenya, Thailand, Myanmar, Vietnam, Philippines), with each study area divided into 18-20 clusters (sectors or neighbourhoods) of approximately 50-100 households each. Using a baseline pupal-demographic survey, the most productive container types were identified which produced >or=55% of all Ae. aegypti pupae. Clusters were then paired based on similar pupae per person indices. One cluster from each pair was randomly allocated to receive the targeted vector control intervention; the other received the 'blanket' (non-targeted) intervention attempting to reach all water holding containers. RESULTS: The pupal-demographic baseline survey showed a large variation of productive container types across all study sites. In four sites the vector control interventions in both study arms were insecticidal and in the other four sites, non-insecticidal (environmental management and/or biological control methods). Both approaches were associated with a reduction of outcome indicators in the targeted and non-targeted intervention arm of the six study sites where the follow up study was conducted (PPI, Pupae per Person Index and BI, Breteau Index). Targeted interventions were as effective as non-targeted ones in terms of PPI. The direct costs per house reached were lower in targeted intervention clusters than in non-targeted intervention clusters with only one exception, where the targeted intervention was delivered through staff-intensive social mobilization. CONCLUSIONS: Targeting only the most productive water container types (roughly half of all water holding container types) was as effective in lowering entomological indices as targeting all water holding containers at lower implementation costs. Further research is required to establish the most efficacious method or combination of methods for targeted dengue vector interventions.

Comparison and critical appraisal of dengue clinical guidelines and their use in Asia and Latin America.

The World Health Organization (WHO) dengue classification scheme for dengue fever (DF) and dengue haemorrhagic fever (DHF)/dengue shock syndrome (DSS) has been adopted as the standard for diagnosis, clinical management and reporting. In recent years, difficulties in applying the WHO case classification have been reported in several countries. A multicenter study was carried out in Asia and Latin America to analyze the variation and utility of dengue clinical guidelines (DCGs) taking as reference the WHO/PAHO guidelines (1994) and the WHO/SEARO guidelines (1998). A document analysis of 13 dengue guidelines was followed by a questionnaire and Focus Group discussions (FGDs) with 858 health care providers in seven countries. Differences in DCGs of the 13 countries were identified including the concept of warning signs, case classification, use of treatment algorithms and grading into levels of severity. The questionnaires and FGDs revealed (1) inaccessibility of DCGs, (2) lack of training, (3) insufficient number of staff to correctly apply the DCGs at the frontline and (4) the unavailability of diagnostic tests. The differences of the DCGs and the inconsistency in their application suggest a need to re-evaluate and standardise DCGs. This applies especially to case classification and case management.

Visceral leishmaniasis on the Indian sub-continent: a multi-centre study of the costs of three interventions for the control of the sandfly vector, Phlebotomus argentipes.

The sandflies that transmit the parasites causing human visceral leishmaniasis (VL) can be controlled by several methods, including indoor residual spraying (IRS), the use of long-lasting insecticidal nets (LLIN) and ecological vector management (EVM). The financial costs of each of these three methods of sandfly control have recently been assessed and compared, in a multi-centre study based on the Indian sub-continent. In each of the four study sites (two in Nepal and one each in India and Bangladesh), 24 neighbourhoods were randomly selected in districts with high incidences of VL. The costs of the three interventions were then prospectively assessed in each study neighbourhood, in the local currency, and then converted to U.S. dollars at the prevailing exchange rate in the country concerned. The costs of IRS, which ranged from U.S.$2.4-11.7 (mean = U.S.$5.9) per household-year, were greater than those of LLIN (U.S.$3.5-5.1/household-year, with a mean of U. S.$4.5) but less than those of EVM (U. S.$5.0- 14.0/household-year, with a mean of U.S.$8.7). These results indicate that LLIN and IRS may be the cheaper options for the control of sandflies on the Indian sub-continent, and that EVM should perhaps only be taken up as a complimentary and voluntary method. Various combinations of these interventions (based on country-specific social and economic factors) may, however, be the best and most cost-effective choice.

Can visceral leishmaniasis be eliminated from Asia?

Data on the burden of visceral leishmaniasis (VL) in Indian sub-continent are vital for elimination programme planners for estimating resource requirements, effective implementation and monitoring of elimination programme. In Indian sub-continent, about 200 million population is at risk of VL. Nearly 25,000-40,000 cases and 200-300 deaths are reported every year, but these are grossly underestimates. Recent well-designed multicentric studies identified VL burden of 21 cases/10,000 among sampled population in Indian sub-continent (Bangladesh, India and Nepal). This estimates 4,20,000 cases per 200 million risk population clearly indicating that the disease is highly under-reported. Chemical and environmental vector control studies show that the indoor residual spraying (IRS) and long-lasting insecticidal nets (LLINs) are effective and significantly reduce sandfly densities. The findings documented from different sources revealed that some gaps and weakness in existing policies for introducing VL vector control interventions. Our studies emphasize the need of integrated vector management with both IRS and LLIN vector control interventions. Active case detection with rK39 strip test as diagnostic tool is the key element for detection of VL cases. The use of oral drug miltefosine for the treatment after assessing feasibility at community level is important. Kala-azar elimination in Indian sub-continent is possible if elimination programmes ensure access to health care and prevention of kala-azar for people at risk with particular attention to the poorest and marginalized groups. The evidence-based policy should be designed that motivates to implement the programmes, which will be cost-effective. Maintaining the acceptable level of incidence requires public awareness, vector control, appropriate diagnosis and treatment. The five pillars of VL elimination strategies identified are: early diagnosis and complete treatment; integrated vector management and vector surveillance; effective disease surveillance through passive and active case detection; social mobilization and building partnerships; and clinical and operational research which need to be re-enforced to effective implementation.

What do community-based dengue control programmes achieve? A systematic review of published evaluations.

Owing to increased epidemic activity and difficulties in controlling the insect vector, dengue has become a major public health problem in many parts of the tropics. The objective of this review is to analyse evidence regarding the achievements of community-based dengue control programmes. Medline, EMBASE, WHOLIS and the Cochrane Database of Systematic Reviews were searched (all to March 2005) to identify potentially relevant articles using keywords such as 'Aedes', 'dengue', 'breeding habits', 'housing' and 'community intervention'. According to the evaluation criteria recommended by the Cochrane Effective Practice and Organisation of Care Review Group, only studies that met the inclusion criteria of randomised controlled trials (RCT), controlled clinical trials (CCT), controlled before and after trials (CBA) or interrupted time series (ITS) were included. Eleven of 1091 studies met the inclusion criteria. Of these, two were RCTs, six were CBAs and three were ITS. The selected studies varied widely with respect to target groups, intervention procedures and outcome measurements. Six studies combined community participation programmes with dengue control tools. Methodological weaknesses were found in all studies: only two papers reported confidence intervals (95% CI); five studies reported P-values; two studies recognised the importance of water container productivity as a measure for vector density; in no study was cluster randomisation attempted; and in no study were costs and sustainability assessed. Evidence that community-based dengue control programmes alone and in combination with other control activities can enhance the effectiveness of dengue control programmes is weak.

Dengue research and training supported through the World Health Organization.

The rapidly increasing burden of dengue, the varied and often poorly understood factors contributing to its global spread, and the challenges of preventing and controlling it have led to a renewed call for more research and training on the disease. The main aims are to improve vector control, case management, and primary prevention through vaccine development. The World Health Organization (WHO), through its inter-departmental working group on dengue, is actively engaged in supporting and co-ordinating the major research activities. The dengue research initiatives of the Special Programme for Research and Training in Tropical Diseases (TDR), other departments at the WHO's Geneva headquarters, the WHO's regional and country offices, and the organization's dengue-affected member states are summarized in this article. This intensified effort, in close collaboration with other stakeholders, is contributing towards the goals of reversing the current epidemiological trends and of reducing the global burden posed by dengue in all of its forms.

Combined field and laboratory evaluation of a long-term impregnated bednet, PermaNet.

For how long does the insecticidal effect of long-lasting nets achieve high Anopheles mortality? Four PermaNets tested previously in Colombia for wash resistance were bioassayed again after 3 years of use and 23 washes. The tests were done both in Colombia and a reference laboratory in Indonesia. Additionally the residual concentration of deltamethrin on the nets was chemically analysed. The results showed high Anopheles mortality and a mean deltamethrin concentration of 9.6 mg/m2. Discrepancies with other studies are discussed.

Economic costs associated with inadequate drug prescribing: an exploratory study in Chiapas, Mexico.

An exploratory study assessing the economic consequences of inadequate prescribing by health-care providers was carried out in two cities of Chiapas State, South Mexico. Two research methods were used: (a) an exit survey with pharmacy users (1190 interviews) and (b) in-depth individual interviews with patients at hospital level (19 interviews). For comparative purposes, three tracer conditions were selected: acute respiratory infection (ARI), diarrhoeal disease (DD) and hypertension. The main findings were that doctors, drug vendors at pharmacies, traditional healers, users of health-care services themselves (through self-prescription) and other health staff are all sources of inadequate prescribing and impose a substantial economic burden of unnecessary cost on health-care users. In general, treatment costs were high compared with standard treatments, and prescriptions included a high proportion of non-essential drugs (50-64%). Additional costs to the standard treatment were particularly high when the prescription was by a medical doctor (US$ 3.57 per ARI prescription and US$ 8.37 for DD). Losses attributable to inadequate prescribing averaged US$ 47 per patient. The study concludes that training programmes on rational drug use aimed exclusively at medical doctors can only have a limited effect in reducing economic losses due to inadequate prescribing. Such efforts should be extended to other providers who are very active in prescribing drugs and providing health advice. The main challenge that the Mexican health system will be facing in the foreseeable future is not only to improve equity of access to essential drugs, but also to improve quality of drug advice and prescription and rational use of drugs by the population.

Unit costs for house spraying and bednet impregnation with residual insecticides in Colombia: a management tool for the control of vector-borne disease.

A study of unit costs and cost components of two malaria-control strategies (house spraying and bednet impregnation with residual insecticides) was undertaken in 11 malaria-endemic states (departamentos) of Colombia, using data provided by control staff on self-administered questionnaires. The accuracy of the data was verified by personal visits, telephone conversations and complementary information from 10 other states. Allthe financial-cost components of the malaria-control operations carried out in the previous 6 months and the results of the control operations themselves (including the numbers of houses sprayed and numbers of bednets impregnated/day) were recorded. The information was stratified according to whether the target communities were 'near' or 'far away' from an operational base, the far-away communities being those that needed overnight stays by the control staff. The main variables analysed were unit costs/house treated, and annual cost/person protected. The results show that house spraying was generally more expensive for the health services than bednet impregnation. This is particularly the case in 'nearby' communities, where most of those at-risk live. In such communities, spraying one house was 7.2 times more expensive than impregnating one bednet. Even if only those sleeping under an impregnated net were assumed to be protected, the unit costs/person protected in a 'nearby' community were twice as high for house spraying than for bednet impregnation. In 'nearby' communities, where technicians could return to the operational base each evening, insecticides made up 80% of the total spraying costs and 42% of the costs of bednet impregnation. In 'far-away' communities, however, salaries and 'per diems' were the most important cost components, representing, respectively, 23% and 22% of the costs of spraying, and 34% plus 27% of the costs of impregnation. Insecticide wastage and non-use of discounts on insecticide prices (available through the national Ministry of Health) increased the overall costs considerably. The multiple uses of these cost calculations for district health managers are presented.

Randomized, controlled, double-blind trial of topical treatment of cutaneous leishmaniasis with paromomycin plus methylbenzethonium chloride ointment in Guatemala.

A double-blind, randomized trial was undertaken in Guatemala to determine the therapeutic efficacy of an ointment for the treatment of cutaneous leishmaniasis that contained 15% paromomycin and 12% methylbenzethonium chloride and that was applied twice a day for 20 days. The treatment group included 35 patients, and the placebo group included 33 patients. The initial clinical response rate (13 weeks after completing the treatment) was 91.4% in the treatment group and 39.4% in the placebo group. The final clinical response rate at the 12-month follow-up examination was 85.7% (31 of 35) in the treatment group and 39.4% (13 of 33) in the placebo group (P < or = 0.001). In general, the treatment was well tolerated and was never interrupted because of adverse effects. The number of adverse effects reported in the placebo group was lower than in the treatment group (16 events versus 30 events). All adverse effects reported by patients disappeared within 1 week of completing the treatment. Our findings show that the combination of paromomycin with methylbenzethonium chloride for 20 days is a good alternative for antimonial treatments of cutaneous leishmaniasis in Guatemala.

Inadequate drug advice in the pharmacies of Guatemala and Mexico: the scale of the problem and explanatory factors.

A study of urban pharmacies in Guatemala and in Chiapas, southern Mexico, was undertaken to analyse the scale of the inadequate drug advice provided, and to identify the contributing factors. The estimate of the scale of the problem was based on the results of 969 approaches to 191 pharmacies by 'magic clients' (i.e. researchers pretending to be looking for treatment for relatives who had one of three 'tracer' diseases), interviews with 4469 pharmacy users as they left the same 191 pharmacies, and a comparison of the quality of advice offered by public and private pharmacies in Guatemala (based on exit interviews with 150 users). The contributing factors were explored using a provider survey (interviews with 166 pharmacy supervisors and 371 drug vendors), an in-depth study analysing large-chain and independent Mexican pharmacies, and a review of the national drug policies in both countries. Although only about 11% of all drug treatments were recommended in pharmacies (the rest being prescribed by physicians or recommended by kin-groups), this still represents large numbers of treatments. Overall, 501 individuals who visited the 191 study pharmacies over 2 days of observation received drugs recommended by pharmacy staff. Much of the pharmacy advice was revealed to be poor: > 80% of the treatments recommended to the 'magic clients' for diarrhoeal disease or acute respiratory infection included unnecessary or dangerous drugs. Few of those who worked in the pharmacies based their advice on careful case histories. Drug advice in pharmacies was much more likely to be of poor quality than that from physicians or even kin-groups. The factors behind this poor advice were identified as a lack of knowledge about standard treatments and legal regulations, incompetence among pharmacy staff, commercial pressures (particularly in the large-chain pharmacies of Mexico), and a failure to implement the existing regulations covering the drug market and its retail practices. It is recommended that: (1) pharmacy owners and drug vendors be made more aware that the selling of drugs should involve provision of healthcare (as well as reasonable profit-making); (2) existing drug-related legislation be reinforced (through consensus-building rather than coercion); and (3) mass training of pharmacy supervisors and drug vendors, in the standard treatment of common diseases, be undertaken. This process will be challenging and slow.