Malaria malpractice

`The Lancet’ charges the World Bank in India with medical malpractice in malaria programmes, but the government is equally culpable.

ON April 25, to mark Africa Malaria Day, the medical journal The Lancet carried a paper that made startling news around the world. “The World Bank: false financial and statistical accounts and medical malpractice in malaria treatment” was the strongly worded title of the paper, which was written by a team of Canadian health experts led by Amir Attaran. It made several serious charges against the Bank’s malaria control programmes in different parts of the world. Notably, the authors charged World Bank in India with knowingly funding ineffective treatment for a potentially fatal strain of malaria against the recommendations of the World Health Organisation (WHO).

India has reported about 2 million malaria cases every year since the mid 1990s. There are two predominant strains: Plasmodium vivax and Plasmodium falciparum. The disease caused by the former is milder, results in death rarely and can be treated with chloroquine. The latter leads to complications, including cerebral malaria. Almost all malaria-related deaths are cases of falciparum; the reason for its deadly nature is that the parasite has become increasingly resistant to chloroquine in almost all malaria endemic regions of the world.

The authors allege that on six occasions in 2004, the Bank approved the purchase of over 100 million tablets of chloroquine, worth $1.8 m, for its project in India knowing full well that the medication would be used to treat drug-resistant falciparum malaria. According to WHO’s revised recommendations of 2003 for treating falciparum malaria, chloroquine should not be used when treatment failure exceeds 15 per cent. “The quantities make it probable that millions of patients with falciparum malaria received such treatment inappropriately. Both money and lives are wasted by these decisions,” the paper said.

In a rejoinder, published in the same issue of the journal, Jean-Louis Sarbib and associates from the Bank rejected the charge of “medical malpractice” and defended the Bank’s Enhanced Malaria Control Project (EMCP) in India. According to them, the Bank’s aim in India, as elsewhere, is to provide credit to the ongoing government-led strategy. This tailored the malaria drug policy to match the distribution of different malaria parasites in various parts of the country. The EMCP covered 1,045 Primary Health Centres (PHCs) in 100 districts over the States of Andhra Pradesh, Chhattisgarh, Gujarat, Jharkhand, Madhya Pradesh, Maharashtra, Orissa and Rajasthan at a cost of about Rs. 120 crores.

Malaria has already made the headlines in Indian newspapers this year. An outbreak in Assam in February, earlier than the usual June-July period, has seen 170 deaths in the State to date, the highest in recent years. Though outbreaks in the northeastern region have now become an annual occurrence, the World Bank controversy has once again brought the government’s malaria control programme, its drug policy in particular, into sharp focus.

India’s total number of malaria cases has declined to a steady 2 million. However, the parasite profile has been changing significantly over the years with a steady increase in the percentage of falciparum cases across the country. According to the Health Ministry’s National Vector Borne Diseases Control Programme (NVBDCP), formerly known as the National Anti Malaria Programme (NAMP), falciparum accounted for 50.56 per cent of the cases in 2005 (see graph) and is present in all parts of the country (see map). Areas with more than 30 per cent of falciparum cases are categorised as high-risk. These include much of the northeastern region (with Assam and Meghalaya over 70 per cent), Orissa (over 80 per cent), Jharkhand, Chattisgarh and even parts of West Bengal, Madhya Pradesh, Maharashtra and Andhra Pradesh.

Only about 0.5 per cent of falciparum cases develop into complicated malaria in the country as a whole. However studies conducted by the National Institute of Malaria Research (NIMR), formerly Malaria Research Centre, in several hospitals in Orissa, Assam and Jharkhand have shown that 25-30 per cent of patients develop complicated malaria. According to its director A.P. Dash, about 30 per cent of complicated falciparum malaria cases become fatal if treatment fails. Annually, Orissa accounts for the largest number of deaths due to malaria in the country. There is also a noticeable increase in the number of complicated falciparum malaria cases and consequent deaths, which indicates of falciparum’s increasing drug resistance, Dash points out. In 2003, the total number of cases was 1.87 million of which falciparum accounted for 0.86 million and there were 1,006 deaths.

But, more pertinently, these reported cases are only indicative of a trend. According to WHO, the actual disease burden is about 6-7 times the reported number. Correspondingly there is a wide gap between estimated deaths and the number actually reported. Any malaria control programme, the drug policy in particular, should take this into account, Dash points out.

According to India’s Malaria Drug Policy, formulated in 2002, chloroquine is the first-line drug for presumptive treatment. It is also the first-line drug for radical treatment of falciparum malaria if the patient is in a `low risk’ area. In `high risk’ areas however, the policy advocates a regimen of chloroquine plus primaquine.

Second-line treatment, a combination of sulphalene/sulphadoxine and pyrimethamine (the SP combination) followed by primaquine, is recommended at the block/PHC level only in `chloroquine resistant’ areas (treatment failure of 25 per cent or more) and in specific cases where a patient does not respond to chloroquine. In severe and complicated falciparum malaria cases, the policy recommends intravenous quinine irrespective of chloroquine resistance status of the area (with a quick switch over to oral quinine) or injectible artemisinine derivatives.

For falciparum malaria, artemisinin-based combination therapies (ACTs) are emerging as the drug of choice for first-line treatment in nearly all other endemic regions of the world. In 2003, the WHO recommended switching to ACT wherever there is a 15 per cent failure rate with chloroquine. A 10-15 per cent failure rate was recommended as an `alert’. However, the Indian malaria control programme did not adopt the changed guidelines on the grounds that reliable countrywide drug resistance data were not available to change the drug policy.

One of the three chief criteria for the selection of blocks in the Bank project was that they should be in a high-risk area. That is, the area should havemore than 30 per cent of falciparum cases and reported deaths due to malaria. In addition, States with low performance and high burden such as Orissa and Jharkhand were chosen. Given this, the treatment policy approved by the Bank for EMCP would seem inappropriate, as it indeed did to the authors of The Lancet paper. But the Bank, it would seem, was merely acting as a fund facilitator, supporting the Indian government’s malaria control policy, which is to administer chloroquine for presumptive treatment.

“The distribution of P. falciparum malaria and patterns of its resistance to chloroquine are not homogeneous [in India],” wrote Sarbib and associates in their response paper. “Therefore, the government’s policy on use of anti-malarial drugs is based on local situation within the framework of an overarching national policy guided by evidence and expert advice. The type of parasite resistance, pattern and cost and availability of drugs are important issues for any changes in drug policy,” they said. These are, in fact, the arguments NVBDCP also advances for the current treatment policy.

The contention of Attaran and associates was that, even if India’s official policy was to give chloroquine as the presumptive or first line treatment, the Bank had “a duty not to supply the wrong treatment to patients and could have financing conditional on India adopting effective malaria treatments”. That is, the paper questioned the Bank effectively endorsing India’s “wrong” official malaria drug treatment protocol and argued that ACT should be used to treat all malaria cases in India.

“The Indian situation cannot be compared with situations elsewhere... The fact that we have been able to control the current outbreak in Assam within 15 days shows that chloroquine is still effective,” said P. L. Joshi, director, NVBDCP. But others disagree. The NIMR has been monitoring the situation on ground continuously; according to Dash, sporadic deaths continue to occur. “With deaths still being reported, you cannot say that the situation has been controlled,” he said.

The WHO guidelines have, in the meanwhile, been changed further. In 2005 the `alert’ phase was removed and ACT was recommended as the first-line treatment from the 10 per cent failure rate. “WHO guidelines are only recommendatory,” said Joshi. “These have to be adopted into our [national] policy, taking into account our field experience across the country, which can vary from one PHC or block to another within the same district.” According to him, field experience has shown that that there are still many regions in the country where falciparum has been found to be sensitive to chloroquine. “The problem lies in effective implementation of all components of the malaria control programme,” he said.

There are many strains of falciparum found in India. “It is possible that some of these strains may be sensitive to chloroquine. But this has to be established through a study, which is lacking,” Dash said. Today only a few countries in Central America north of the Panama Canal, including Haiti and Dominican Republic, do not report chloroquine-resistant falciparum malaria.

Nick J. White, Chairman of the WHO committee responsible for advice on antimalarials, has challenged the Bank’s position. “WHO guidelines for the treatment of malaria make it clear that there are very few places in the world where chloroquine can be used to treat falciparum malaria and India is not one of them,” he wrote in a note intended for Sarbib, which is now in the public domain. “Chloroquine resistance in P. falciparum has fuelled the resurgence of malaria in India over the past two decades. The only situations where it would be justified to deploy chloroquine for the treatment of suspected malaria are in areas where only P. vivax is present or where there is unequivocal evidence of chloroquine susceptibility based on 28-day in vivo tests supported by in vitro tests and molecular markers. I cannot find evidence to support your contention that chloroquine still works for falciparum malaria in areas where the Bank has supported chloroquine deployment,” he said.

Joshi has contended that the gradual introduction of ACT was preferable to a nationwide changeover to avoid the parasite quickly developing resistance to ACT. This argument is somewhat far-fetched because the very idea of combination therapy is to prevent drug resistance. “In principle, this is possible but is extremely improbable,” pointed out Dash.

As is evident from the above, essential to an effective drug policy is a proper nationwide surveillance of chloroquine resistance and a realistic mapping of `chloroquine-resistant’ areas in the country. In India, chloroquine-resistant falciparum malaria was first identified in Karbi Anglong in Assam in 1973. Since 1978, drug resistance is being monitored by NAMP/NVBDCP through 13 monitoring teams, each comprising eight technicians, located in 11 regional offices countrywide. This number may have been adequate when monitoring first began nearly 30 years ago. But, inexplicably, it has not increased. Given that chloroquine resistance has become quite widespread, with pockets showing 90 per cent resistance according to some studies, monitoring should be intensified.

The NVBDCP has to date identified 241 PHCs in 57 districts as `chloroquine resistant’ areas where treatment failure has exceeded 25 per cent. In all, these PHCs, second-line treatment of the SP combination followed by primaquine is being followed with the recommended changeover to ACT yet to be implemented in all of them. According to NVBDCP, 12 PHCs in seven States have already reported various levels of resistance to the SP combination. It is quite likely, therefore, that any ACT pilot phase would begin with these PHCs. In fact, 21 of these 241 PHCs come under the Bank’s EMCP. According to G. N.V. Ramana of the World Bank in New Delhi coordinating EMCP, the project has provided ACT packs in these areas. “We had approved 3 million doses but only 0.7 million could be procured because two bidders could not supply given the high global demand,” he said.

The identification of 241 PHCs as `chloroquine-resistant’, however, needs to be qualified. First, this identification is not based on a rigorous countrywide surveillance programme. Basically, these were the PHCs studied by the monitoring teams. “Wherever surveillance has been carried out, resistance has been found,” points out Dash. “If you expand and intensify the surveillance, you will discover more.” Second the true resistance pattern could be masked because identification is based on an earlier surveillance protocol.

Until 2001, the protocol followed by NAMP/NVBDCP was based on 7-day follow-up studies using in vivo tests for parasite response to the drug. Infections were classified as drug-sensitive (S) drug-resistant (R) at one of three levels, RI, RII or RIII, with RIII being the most resistant. RI responses represent an initial clearance of parasitaemia followed by recrudescence eight or more days after the treatment; RII responses represent the clearance or substantial reduction of parasitaemia with recrudescence on day 7 and RIII refers to a situation where there is no initial reduction and the parasitaemia may even increase. In a 7-day follow-up protocol, it will be difficult to distinguish RI from S and this level of resistance could be mistakenly classified as Sensitive. Most of the 241 PHCs identified as chloroquine-resistant were classified as such on the basis of the 7-day protocol; of 16,833 falciparum cases tested for drug resistance, 77.5 per cent were classified sensitive to chloroquine, 14 per cent resistant at RI level, 4.8 per cent at RII level and 3.7 per cent at RIII level.

A 28-day follow up Therapeutic Efficacy Study is now used by the NVBDCP for surveying drug resistance, which also will pick up the prevalence of resistance at RI level. Also, the classification of resistance is now based on three categories as under the WHO-modified criteria: Adequate Clinical and Parasitological response (ACPR), Early Treatment Failure (ETF) and Late Treatment Failure (LTF). Drug policy has been changed at PHCs that report 25 per cent or more of total treatment failure (ETF + LTF). It is important to note that ACPR is a measure not of parasitaemia clearance alone but also resolution of clinical symptoms such as acute fever. The revised 28-day protocol will give a truer picture of chloroquine resistance. However, systematic countrywide surveillance on this basis is yet to be carried out.

Since 2002, only 1,551 falciparum cases have been examined according to the revised protocol. Of this, 61.12 per cent have been found showing ACPR and 38.8 per cent have shown total treatment failure. Also, 133 cases were studied for resistance to SP combination, of which 122 (91.73 per cent) showed ACPR. It is on the basis of these data that the NVBDCP has claimed that Indian experience in the field indicates that falciparum malaria is still found to be sensitive in many parts of the country. However in terms of sample sizes these are far too small to give an accurate statistical picture of the drug resistance profile across the country. Furthermore this does not seem to correspond with the extent of genetic mutations responsible for drug resistance that has been found across the country.

While the NVBDCP has only recently put in place a programme of sentinel surveillance to map the true resistance profile, the studies carried out by the NIMR in low transmission areas such as Rajasthan do indicate chloroquine resistance in these areas as well. As can be seen from the maps, the revised protocol shows that chloroquine resistance is more widespread and includes many more low transmission areas than the 57 districts currently listed by the NVBDCP.

Based on such studies, the expert group on Malaria drug policy recommended in October 2005 a change over to second-line ACT (based on SP along with artesunate) in chloroquine resistant areas, following the WHO guidelines. However, there appear to be problems of procurement. There are no domestic manufacturers of the recommended drug, though there are some manufacturers of variants that are currently undertrial, so they must be imported. Cost is another significant factor. ACT costs about 20 times the price of chloroquine, which is about Rs.4.50 a dose. Moreover, according to Joshi, there are also problems of logistics of delivery and implementation down to the PHC level in the Indian context. So the strategy that has been adopted is to pilot the introduction of ACT in a few blocks from the identified chloroquine-resistant areas.

“The process of change in this country is slow,” laments Dash. “PHC-level data generation will take far too long,” he adds. “The sentinel surveillance that has been put in place should generate adequate data quickly and we will switch to ACT wherever required,” asserts Joshi, on the other hand. In the meanwhile, with the completion of EMCP in December 2005, the World Bank has issued an aide-memoir for an external review of the entire project. “This should be completed by December,” says Ramana.

A new Bank-funded Vector-borne Disease Control Programme is due to be operationalised by NVBDCP from 2006. The criticism of the Bank’s malaria programme in India, however, seems to have had an impact. “The new programme will go as a package including use of rapid diagnostic techniques and change in treatment protocol,” says Ramana. Any change in malaria drug policy is, however, not discernible from the NVBDCP’s responses to the situation on the ground. In the meantime, the malaria disease burden, increasing drug resistance and consequent deaths will continue to affect millions in the country.