Literature search

Systematic searches were conducted in MEDLINE, CINAHL, Scopus and the Cochrane Database for Systematic Reviews in October 2019 using a combination of search terms relating to active case detection in concert with the two review diseases. Search results were limited to English language papers from 1999 to 2019. See S1 Appendix for a full list of search terms and results. All search results were exported into EndNote for processing and screening.

Inclusion and exclusion criteria

To be included in the review studies had to examine an ACD campaign, with one of the two review diseases, leprosy and crusted scabies, as the sole or primary campaign target. Included studies needed to report outcome data on the detection rates of the campaign and a relevant comparison such as local prevalence rate (PR), or the detection rate of a concurrently conducted detection method. These criteria led to the exclusion of papers in which ACD is conducted as part of a control program but the ACD activity is not evaluated against another case detection method and outcome.

Data extraction, summary and risk of bias assessment

Data were extracted on the ACD campaign setting (community characteristics and country), ACD strategy type, campaign time frame, personnel, method and use of laboratory evidence (typically skin smear) in diagnosis (if not reported recorded as ‘no’). Outcome data extracted were detection and/or prevalence rates (both campaign and comparison). Disease stage at diagnosis was not sought. Due to heterogeneity in campaign type, reporting and setting, a meta-analysis was not performed.

The relevance of the comparative detection or prevalence rate to the study setting impacts the risk of bias in assessment of outcome effect. A significant difference between the campaign and comparator rates will confirm a positive (or negative) effect of ACD in general. However, if the comparator rate has low relevance, the difference may conflate variations in prevalence with campaign effectiveness. For example, comparing the detection rate of a campaign conducted in a high prevalence district to the national prevalence rate in a low prevalence country will inflate the campaign effect size. To assess this risk of bias at the study level, a grading system was developed to rate the relevance of the comparison rate; the grades are identified in the data table (Table 1) and are:

• Low: location and format of low relevance (e.g. national prevalence rate (PR) to campaign new case detection rate (NCDR) in regional sample)
• Moderate: location relevant but comparator rate format and/or timeframe not relevant (e.g. local PR rate to campaign NCDR)
• High: location, timeframe and comparator rate format relevant (e.g. NCDRs of campaigns conducted concurrently and/or in same region).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Data summary.

https://doi.org/10.1371/journal.pntd.0009577.t001

It is important to note that assessment against the comparator rate may not be the objective of included studies, but may rather serve as context. There is no evidence that studies examining a single campaign selectively report less relevant comparator rates; when these are used it is assumed to reflect the availability of prevalence data.

Study characteristics

All included studies report on ACD campaigns for leprosy. The search yielded only four studies including the term “crusted scabies”, but none met the inclusion criteria. The papers included in the review draw on studies from nine countries, with five from India and three from Brazil (the two countries with largest leprosy numbers globally [17] and the remaining mostly from Asia and Africa). There are no studies from developed countries. Community screening is the most commonly reported detection method, used solely or in concert with contract tracing. Three studies examine contact tracing only.

Not all targeted data extraction points were available in all studies, and extracted data was not uniform. When only detection numbers were reported, the new case detection rate (NCDR) or prevalence rate (PR) (whichever appropriate) was calculated manually. A summary of the included studies is presented in Table 1. Heterogeneity in reporting of both campaign implementation and outcome data inhibit standardization in summation for this review.

Detection methods

Twelve of the reviewed studies examined community screening methods targeting underserved communities and/or endemic regions. The most commonly reported method of community screening used door-to-door (door-to-door) visits by a small team of community health workers (CHWs) for information and education communication plus screening for clinical signs of leprosy, and referral of positive screened (universally referred to as ‘suspects’) to a local health centre for diagnosis. A number of studies reported alternate settings, including schools and village squares.

Six studies examined contact tracing campaigns, three of which used this method to identify high prevalence areas for targeted community screening. One study used the contact tracing for micro-targeting of geographies for screening, which is achieved by identifying and tracing index patients, then mapping their house location and using case clustering for highly localised community screening [31]. In the included studies, contact tracing was retrospective–identifying index patients through historical national and/or notifiable disease records and seeking out both the index and their contacts to either the household level, and in most cases, the neighbour level [30,31,36].

Less than half of the studies used laboratory testing for diagnosis. The WHO guidance for control leprosy in endemic countries states that diagnosis can be made on the presence of a skin lesion consistent with leprosy with definite sensory loss or a positive skin smear [42]. In non-endemic countries, laboratory evidence through skin smear [22] or nucleic acid testing [23] is required. The only laboratory testing reported used in review studies is a slit skin smear sample for microscopy for leprosy bacilli–utilised in half the studies [30,33,39,38].

Detection personnel

Across studies, screening was typically conducted by a small team (2–4) of community health workers and/or local volunteers/workers. Few studies provide any meaningful detail about the recruitment, training or remuneration (if any) of field or community workers/volunteers, or about roles and responsibilities during the campaign (i.e. information/education, skin examination). When the duration of training is reported, it is typically 1–3 days [e.g.37,39]. Most studies report whether the gender composition of the team is mixed or single sex. Female community health workers are used in some settings for cultural appropriateness [e.g.33,37,39]. One study reported using local recovered leprosy patients in community screening as a means to enhance community buy-in [33].

Numerous studies reported making contact with community leaders prior to commencing a community based campaign to gain support and raise awareness [18,39,41]. This practice was most common in rural and village based community settings. In a few studies, the relationship (e.g. prior or on-going contact) between mobile health workers and the campaign community is reported [18,41] A minority of studies report whether consent was sought before entering a household or conducting a skin examination. Those that do tend to be more recent, and often involve a developed country implementation partner [e.g.19,39,41].

Campaign effectiveness

All ACD campaigns in the review studies found new leprosy cases–confirming that routine methods were missing active cases in the sample populations. Heterogeneity in both detection and comparator measures inhibits a meta-analysis of outcomes. Similarly, heterogeneity in the scale, setting and personnel of the reviewed campaigns inhibits a systematic comparative assessment of method design or effectiveness across studies. Data synthesis shows that the highest campaign NCDRs/PRs are in contact tracing campaigns and those conducting community screening of marginalised ethnic groups [18,32,38].

Four studies assessed financial costs. Ezenduka et al [20] compared three ACD methods in Nigeria: household contact tracing, targeted community screening, and a traditional healer incentive to encourage referral to local health centres. The authors found that household contact tracing has the lowest cost per new case detected at US$142/NCD compared to US$192/NCD in the traditional healer incentive and US$313/NCD for community screening. In Tiendrebéogo et al’s [40] study, a community screening ACD campaign (US$72/NCD) cost twice as much per new case detected than PCD (US$36) but yielded a four times higher prevalence rate, and detected cases at an earlier disease stage (the costs/benefits of which are not quantified). Gillini et al [33], reporting on a campaign in Nepal, found a dramatically higher ACD cost than the African studies of US$534/NCD more than the passive method. The baseline cost of the passive method was not reported so total cost is not identifiable. The African studies used local personnel while the Nepalese study involved personnel from WHO and a Japanese non-profit which likely contributed to cost differences. Ganapati et al [33] reported the cost per NCD for a case diagnosed through clinical examination (US$20) was less than 10% of the cost of case diagnosed through skin smear (US$322).

Data synthesis suggest that rapid screening campaigns (<1 month) tend to yield lower rates of suspect confirmation, and find fewer cases than non-rapid methods (typically 6–18 months). When comparing their outcome data with an ACD campaign of similar design and scale but conducted over a much longer time frame (150 days compared with 6 days), Rao et al [37] found that the longer campaign had a NCDR more than twice as high, suggesting rapid survey methods may be less effective. Rapid community screening campaigns were also associated with lower suspect confirmation rates, which is suggestive of high false positive screenings. Seven of the community screening studies, particularly the Indian ones, report both the numbers of positive screened individuals (‘suspects’) and numbers diagnosed. The proportion of suspects confirmed range from 4%-100%, with the average around 10–15%. In the study reporting 4%, a chaser team of medical officers visited the homes of ‘suspects’ identified by community health workers/volunteers meaning 96% of positive screens were false positives. These findings illustrate poor diagnostic capacity amongst community screeners or a low specificity screening test.

In studies relying on ‘suspect’ self-report to local health centres for diagnosis, the proportion of ‘suspects’ who actually attended the health centre is rarely reported. In the few studies that do, attendance rates are roughly 50%. For example, in partial records from the Nepal campaign just under 50% of ‘suspects’ for which this data was captured attended health clinic [33]. In these cases, the difference between suspect numbers and confirmed cases is partly a product of false positives, but also suggestive of barriers to access. The studies with the highest suspect confirmation rates were of longer duration, used more experienced personnel in screening and had campaigns more integrated with local health services [18,39].

Significantly, only two studies reported treatment outcomes. One reported that 74% of cases had completed MDT one-year post campaign [35] and the other reported that 90% had commenced MDT by the end of the six month campaign period [18].

Barriers and enablers to campaign implementation

Five studies reported barriers to and enablers of ACD campaign implementation. De Souza et al [31] found geographic information systems (GIS) an enabler in a context with few traditional address markers (e.g. street sign/number) in Brazil. Mangeard-Lourme et al [18] found evidence of micro-clustering of leprosy cases in a district in India pointing to the value of geo-mapping for resource allocation/campaign targeting. Numerous studies reported the involvement of community volunteers as an enabler in gaining community support for the campaign [37], including former leprosy patients in one study [33]. However, the use of community workers/volunteers may represent a barrier to campaign effectiveness given low screening accuracy rates.

Lack of transport access, inadequate time for comprehensive screening and long waiting times at the local health centre were reported barriers for both patients and community health workers that led to incomplete coverage of households and attendance of suspects in the Nepalese campaign [33]. One study in India reported that the co-occurrence of numerous chronic skin ailments with leprosy was inhibiting proper diagnosis [39].

Two studies reported findings that evidence how low awareness would inhibit PCD effectiveness; in one Indian study, 45% of newly detected cases had visited a health centre in the past 1 to 2 years, most of which had done so for examinations of lesion(s) specifically [39]. This illustrates poor diagnostic capability in local health services. Another Indian study found that only 34% of newly detected cases reported having noticed their skin lesion(s) prior to diagnosis demonstrating low community awareness which would impede patient self-reporting [35]. Non-availability of MDTs a common barrier to treatment completion [39], and has the capacity to discourage self-reporting if patients do not believe diagnosis will result in effective treatment.

Implications for crusted scabies case detection and review limitations

In summary, the potential impact of ACD on crusted scabies is not quantified from review findings, however, findings show that ACD campaigns tend to find the most missed cases of leprosy in geographically isolated and culturally marginalized communities, which aligns with evidence about ACD more generally [13,14]. This suggests an ACD campaign for crusted scabies in Australia may be worthwhile, given the marginalisation of remote Aboriginal communities and significant barriers to healthcare access amongst this population. The cost effectiveness, particularly for community-wide screening, is not clear given the dearth of cost studies and low applicability of existing ones to the NT. Future cost effectiveness studies in Australia should assess ACD for both simple and crusted scabies in regions where both are endemic. As all studies reported in this review are in developing country contexts and communities with significantly higher population densities than remote northern Australia the transferability of particular models is limited. However, the review illustrates a number of findings that may inform ACD program design in Australia.

As crusted scabies is now a notifiable disease in the NT, contact tracing is a mandated component of disease control making some of the findings from contact tracing campaigns for leprosy relevant. At present, contact tracing is conducted to the household level for new cases of crusted scabies to identify and treat any cases of scabies, both to support scabies control and to ensure the index case does not return to a scabies endemic environment. This review points to the potential value of tracing beyond the household, with cases often found at the neighbour level to index cases in leprosy–a disease less infectious than crusted scabies. This would also align with the high levels of inter-household mobility common in remote Aboriginal communities [44].

The use of historical disease records for identifying index cases may also be of value for ACD in crusted scabies. Similarly to review studies finding historic index cases with active leprosy, crusted scabies cases are often reported as recurrences or on-going due to incomplete treatment [3,10]. Drawing index cases from historical disease records may be a useful starting point for crusted scabies ACD campaign targeting. GIS mapping may also be useful for micro-targeted community screening given the very low population density and the absence of traditional address systems/markers in some remote Aboriginal communities.

A limitation of the literature covered in this review for the Australian context is the reliance on the household as the campaign target. Household centric strategies will miss the most hard-to-reach individuals: the homeless and highly mobile. Recent research found that a high proportion of crusted scabies patients are homeless [10,11]. Any ACD campaign designed for the NT would require consideration of how to capture individuals outside a household system.

There is limited information from which to draw conclusions about the likely acceptability of review ACD programs in Aboriginal communities. Only a few studies discuss the process of consent–an issue that would be highly important in the Australian context. Cultural sensitivity was supported in some studies through the use of community workers/volunteers and making contact with village elders/leaders to gain support prior to community outreach. The involvement of recovered patients is a novel and potentially valuable concept. There is a small but important and highly relevant grey literature about leprosy control in remote Aboriginal communities by the primary clinician operating in the area in the peak endemic period from the 1950s-70s, Dr John Hargrave [27,28,45]. This literature includes a field guide for Aboriginal health workers, whom Hargrave viewed as integral to leprosy control, and discussion of the role of culture in leprosy management, and relationships between communities and healthcare providers.

The review finding of poor screening accuracy by community health workers/volunteers resulting in high numbers of false positives highlights the importance of diagnostic skilling in ACD teams. Lack of healthcare provider awareness and diagnostic skills in low prevalence areas is known to inhibit early and accurate detection of both leprosy and crusted scabies [10,22,46]. There is limited evidence from this review about how diagnostic up-skilling, or laboratory testing and associated patient re-tracing can be managed in leprosy ACDs as only a minority of studies reported it.

The role of new diagnostics needs consideration for both leprosy and scabies ACD. Molecular detection using nucleic acid amplification techniques such as polymerase chain reaction (PCR) to detect M. leprae has been used on skin and nasal swabs [47] and also shows promise for diagnosis of scabies [48]. Point of care diagnostics for scabies are currently under development [49]. As they target detection of scabies mite DNA, these tests are likely to be highly sensitive for crusted scabies and thereby assist health workers in the field with diagnosis of individuals with suspected crusted scabies. Sensitivity is less in ordinary scabies with the much lower mite burden, so a role of POC diagnostic for ACD for scabies in general remains to be determined. Recent innovations in non-invasive diagnostic techniques such as video dermatoscopy [50] may also aide the accuracy of community based screening in the identification of crusted scabies. Increasing use of teledermatology that may also be of benefit to crusted scabies ACD, with even basic photography by relatives being used to support crusted scabies identification in nursing home patients [51] and specialist review of field workers’ photographs of ‘suspect’ crusted scabies cases proving effective in MDA programs [7]. Further research is required to assess the effectiveness of these technological advances in ACD programs.