Communities of practice
The Impact Of Multidrug Therapy On Trends In Transmission
1 Nov 2007
Source: WHO/TDR
Abraham Meima
Erasmus University Rotterdam, Department of Public Health Medical Decision Sciences, Rotterdam, The Netherlands
Working paper for the Scientific Working Group meeting on Leprosy Research, convened by the Special Programme for Research and Training in Tropical Diseases, Geneva, 26–28 February 2002
Full text source: Scientific Working Group, Report on Leprosy, 26–28 November 2002, Geneva, Swizterland, Copyright © World Health Organziation on behalf of the Special Programme for Research and Training in Tropical Diseases, 2003, http://www.who.int/tdr/publications/publications/swg_leprosy.htm 
Literature review of trends in leprosy new case detection rates
As a first step in investigating the impact of case detection and multidrug treatment (MDT) on leprosy, a literature review was conducted of trends in leprosy new case detection rates (NCDRs) as published in international literature [1]. This review covered nine countries and seven smaller geographical entities,a and NCDR data up to 1993. For the majority of the areas/countries, NCDR trends were declining (13 of 16 areas/countries had an average annual decline in NCDR of at least 2% per year). The literature review may have suffered from publication bias, i.e. a tendency to publish papers on high quality control programmes covering long periods of time or indicating successful leprosy control. In the review, an acceleration of declines in trends after the introduction of MDT was not visible. The long incubation period of leprosy could have masked such accelerations. For seven of the nine countries, additional country data became available [2–7], which allowed for extension of the NCDR time series. A general impact of MDT on NCDR trends can still not be demonstrated.
New case detection rate time series constructed on the basis of country data for 1985–2001
For 14 countries with at least 2000 newly detected cases in 1998,b NCDR times series could be constructed from the year 1985 onwards using country data obtained from subsequent issues of the Weekly Epidemiological Record [3–7] and from one conference report [2]. The NCDR trends are highly variable, and underlying trends in leprosy transmission are unclear. Most probably, operational factors heavily influenced the NCDR trends: leprosy control activities were intensified following the 1991 World Health Assembly resolution to “eliminate leprosy as a public health problem by the year 2000”, and leprosy elimination campaigns were initiated. For 7 of 14 countries, NCDRs in the 1990s were either rather stable, or increased (Bangladesh, Brazil, Ethiopia, India, Indonesia, Mozambique, Sudan). For 3 of 14 countries, sudden sharp increases in NCDRs were observed in the late 1990s (Myanmar, Nepal, Madagascar). For 3 of 14 countries, a decreasing tendency - at least in more recent years - was observed (Philippines, China, Vietnam). For the remaining country (Guinea), strong fluctuations in the NCDR were observed.
Reasons for declining trends in the transmission and incidence of leprosy
Declines in transmission and incidence (i.e. onset of disease) of leprosy may be related to several factors:
The period during which M. leprae is transmitted, which can be reduced by early case detection and chemotherapy treatment.
BCG vaccination, which is widely administered as a preventive measure against tuberculosis but appears to afford more protection against leprosy than tuberculosis [8].
Socioeconomic conditions, which are thought to play an important role in leprosy [9]. Their improvement may result in a decline in incidence. Factors suggested to contribute are housing conditions, number of persons per household per room, family size and nutritional factors.
Possible protection of tuberculosis against leprosy [10], either by immunization or by competing risk.
Factors possibly limiting the impact of bcg
In most developing countries, BCG is given in young childhood, and is only given once. Protective efficacies against leprosy ranged from 20% to 80% in randomized controlled trials [8]. But the protection was quite small in Asia where most patients are detected (two trials in India: 24% and 34% [11]; one trial in Burma: 20%). It is also unclear whether the protection decreases with age. Thus, the impact of BCG on leprosy trends may be smaller than one would hope for. The impact also depends on the extent of coverage by BCG vaccination, which in many countries was not high before the 1990s (WHO disseminates country data regarding coverage of immunization programmes through http://www.who.int/vaccines-surveillance/intro.html ).
Factors possibly limiting the impact of MDT based control
The assumption that case detection and treatment would reduce leprosy transmission is reasonable, but the reality may be more complicated. Individuals incubating the disease may already harbour many bacilli, and it is possible that these individuals already transmit M. leprae to others long before the onset of disease, given its long incubation period. Such transmission cannot be prevented by early detection and treatment.
A further problem is the delay between onset of disease and detection. For instance, in the ALERT control programme in Ethiopia, the average detection delay exceeded two years. How easily leprosy is transmitted is not known. The group at risk of developing leprosy might be small, possibly due to genetic factors (leprosy infection is suggested to be much more common than leprosy disease [10,12]) or because close contact is important. Close contact - household and family, neighbours, social and business contact - has been suggested to play a key role in transmission [13]. It is well possible that close contacts of a leprosy patient become infected rapidly. If close contact is indeed important, this may lead to a rapid decrease in the patient's opportunities to transmit M. leprae. Thus, “early” detection may still be too late to prevent much of transmission by subsequent treatment. Other factors which could limit the impact of leprosy control have also been suggested, including carriage of M. leprae in the nose, persistence of M. leprae in the soil, and even animal reservoirs [14–17].
Leprosy simulation model
An epidemiological model, SIMLEP, was developed in order to assess the impact of interventions on transmission. In this model, populations are divided in mutually exclusive compartments, and a set of calculation rules is used to determine the number of transitions between the compartments in subsequent time steps [18]. Through this approach, trends over time can be simulated in transmission and onset of disease, case detection, numbers of incubating individuals, undetected leprosy patients, and patients on treatment.
Analysis of disappearance of leprosy from norway using SIMLEP
Leprosy was still an endemic disease in Norway around 1850, but had virtually disappeared by 1920, long before effective anti-leprosy treatment became available. The downward trend is extremely well documented [19]. The decline coincided with continuous growth of the Norwegian economy. In Norway, a policy of isolation of patients was implemented. By legislation in 1877 and 1885, leprosy patients either had to be isolated in separate rooms in their houses, or had to be admitted to a hospital.
The downward trend of new case detection in Norway was adequately reproduced with SIMLEP, and equally well for 8 (2 × 4) pairs of assumptions on contagiousness during the incubation period and on decrease of transmission opportunities (build-up of contagiousness during incubation period: yes/no; half-value time for transmission opportunities: “none”, 2, 4 and 8 years). However, the estimated contribution of hospital isolation to the decline of leprosy in Norway ranged from only 3% to 60% for these 8 pairs of assumptions, the other explanation being the socioeconomic development. Thus, the impact of isolation proved to be very uncertain.
“Natural experiment”
The Norway study showed that declining trends in leprosy transmission may have competing explanations. A logical next step in a model-based approach would be a SIMLEP-based analysis of long-term trends in (preferably adjacent) geographical areas with comparable general conditions, but with different well documented leprosy control policies. Such trend data would enable disentanglement of competing explanations (MDT, BCG, socioeconomic change) for decreases in transmission, but unfortunately they are not readily available. So far, we have not identified suitable datasets.
Scenario analysis of leprosy trends up to 2020 using SIMLEP
The effect of MDT is similar to that of isolation: both prevent leprosy transmission. Trend predictions are complicated by the same factors that were encountered in the evaluation of the Norwegian decline. Building on the Norwegian experience, a scenario analysis of future trends in leprosy incidence was conducted with SIMLEP [20].
For each of the 8 pairs of assumptions on contagiousness during the incubation period and on decrease of transmission opportunities, model projections were fitted to reference data on leprosy new case detection from 1985 onwards (the reference case detection rate was calculated as the average of the NCDR of the 14 countries that detected at least 2000 cases in 1998, and for which data are available at country level throughout 1985–1998). In doing so, it was assumed that the delay in case detection decreased in the 1990s, which is in line with the intensification of control efforts that took place in many countries. Subsequently, incidence rates were predicted up to the year 2020, assuming an average detection delay of 2 years from 2000 onwards. Again, wide variation was observed in the projections: the annual decline in incidence rate between 2000 and 2020 ranged from 2% to 8% per year for those pairs of assumptions for which a reasonable fit of the reference data was obtained. The corresponding times to reduce the incidence rate by 50% were 43 and 8 years. The rates of decline were lower with contagiousness during the incubation period, and when a faster decrease in transmission opportunities for patients was assumed.
So far, there was no BCG vaccination. The scenario analysis was repeated while making quite favourable assumptions on BCG: 50% lifelong protective efficacy, and rather optimistic coverages. Now, the annual decline in incidence rate ranged from 5% to 10% per year, with corresponding times to reduce the incidence rate by 50% of 14 and 7 years. Thus, leprosy incidence declined in all scenarios considered, and BCG enhanced the declines. The most important conclusion is the slow pace at which the incidence is reduced in all scenarios.
Conclusions
MDT based control appears to reduce transmission. The pace of reduction is highly uncertain, but in any case slow. BCG may enhance the pace, but its impact is also uncertain.
The impact of MDT based control is highly uncertain because of the following unknowns:
the role of close contact in transmission
the speed of transmission
whether, and to what extent, contagiousness builds up during the incubation period.
Research addressing these questions is essential to narrow down the uncertainty regarding the impact of MDT based control.
Further progress through SIMLEP requires data which allow for disentanglement of competing explanations (MDT, BCG, socioeconomic change) for downward trends in leprosy transmission. This requires the availability of long-term trend data from geographical areas with comparable general conditions, but with different well documented leprosy control policies.
Footnotes
b. The 14 countries are: Bangladesh, Brazil, China, Ethiopia, Guinea, India, Indonesia, Madagascar, Mozambique, Myanmar, Nepal, Philippines, Sudan, Vietnam.
References
Comments
Featured Meetings:
Is your organisation working against the infectious diseases of poverty?
There are no comments about this article: Please login if you want to submit a comment.