Search tips
Search criteria 


Logo of brjopthalBritish Journal of OphthalmologyVisit this articleSubmit a manuscriptReceive email alertsContact usBMJ
Br J Ophthalmol. 2007 August; 91(8): 989.
PMCID: PMC1954824

Pragmatism versus purity: effectiveness of the key informant methodology in a developing rural setting

The goal of the VISION2020 initiative is to eliminate avoidable blindness by the year 2020.1 The estimated prevalence of blindness and visual impairment in 2002 was lower than previously predicted (37 million instead of 52 million), suggesting that the three components of VISION2020 – disease control, human resource development and infrastructure development, have successfully helped millions of people escape visual impairment or blindness.2 However, the WHO point out that with an ageing population in many countries, the risk of visual impairment increases and therefore these early successes need to be developed and expanded to meet this challenge.

At the other end of the age range, an example of concrete progress in this worldwide effort is provided by Muhit et al. (see pages 995 and 1000) in this month's BJO. The authors describe a method to add to the existing techniques to ascertain blind children in a developing country. Whereas many valuable studies have used data from children in schools for the blind to estimate the proportion due to different causes, this paper illustrates that important additional information may be obtained by also including children identified by local members of the community, who have been briefed by the study team on what to look for. These individuals, known as “key informants” (KIs) were unpaid volunteers who, after attending a half‐day training session, spent 2–3 weeks in their communities actively seeking out blind or visually impaired children and encouraging their families to bring them to the eye examination carried out by the project team.

The first point to note is that the number of children ascertained by the KIs was nearly double that of the numbers recruited from special and integrated schools or community‐based rehabilitation (CBR) programmes combined (1245, vs. 394 and 296, respectively). Second, the children recruited by the KIs were more likely to have severe visual impairment (SVI) rather than blindness; 9.8% SVI vs. 4.8% in blind schools and 7.8% in community programmes. Third, the cause of their vision loss was 40% more likely to be avoidable than for children identified in schools for the blind and, similarly, 30% more likely to be avoidable than for those identified in CBR programmes. Fourth, the children identified by the KIs were more likely to be female, aged 0–5 years, to live in rural areas and have infantile‐onset eye problems, than children identified by the other two methods – groups which would otherwise have been under‐represented.

The results described by Muhit et al. illustrate the value of this method in the search for accurate prevalence data, particularly for individuals who may benefit from treatment or who might have benefited from preventive strategies. Although this approach has limitations – for example, it is most suited to a society that is relatively open and in which people know others that are in their own geographic area, thereby being limited by transport rather than social convention – it is a valuable addition to existing techniques in population‐based ophthalmological research. The “Gold Standard” of enumerating each person in households in geographically defined clusters and examining them all has been successfully used for common eye conditions in children,3 but the large sample sizes needed for rare conditions (such as blindness in children) make formal population‐based surveys prohibitively expensive and logistically challenging. Even when identifying the population of interest from centralised databases and offering them transport to attend a central facility, it is difficult to achieve high enough compliance rates to provide accurate prevalence data and the databases may not include all individuals of interest. Thus, the use of KIs is a valuable addition to available methods for future studies on the prevalence of blindness and has the potential to make such studies more representative of the whole population. The additional cost of the KI component was only 25% of the whole study, therefore the additional ‘cost per case identified' using the KI method was around a fifth of the ‘cost per case identified' using traditional methodology. The marginal costs associated with the use of KIs were thus very small considering the large number of extra children identified, and the added value of these data to the statistical power and representativeness of the study as a whole. Compared with the cost of carrying out conventional population‐based research in established market economies such as the UK, the value for money represented by this method is impressive and may have applications in developed as well as developing countries.

Without quality population‐based data on prevalence and causes of vision loss, control strategies cannot be devised, nor their effectiveness assessed. Improved data acquisition by this approach will strengthen confidence when making judgements of effectiveness and guide future policies towards better blindness prevention programmes.


Competing interests: None declared.


3. Negrel A D, Maul E, Pokharel G P. et al Refractive error study in children: sampling and measurement methods for a multi‐country survey. Am J Ophthalmol 2000. 129421–426.426 [PubMed]

Articles from The British Journal of Ophthalmology are provided here courtesy of BMJ Publishing Group