Leprosy Mailing List – June 20th, 2012
Ref.: The Diagnosis of Leprosy. One or two cardinal signs?
From: W van Brakel, Amsterdam, The Netherlands
Dear Salvatore,
I would like to respond to the discussion of Dr Manglani and Dr Naafs regarding cardinal signs for the diagnosis of leprosy. This is basically a discussion of what would be the most suitable cut-off point for the diagnosis 'leprosy' – one, two or three cardinal signs. Each of these choices will represent a certain combination of sensitivity and specificity and, also depending on the local prevalence of leprosy, a predictive value of the diagnosis based on the chosen cut-off. The reliability of each individual sign on its own also plays an important role, with nerve palpation probably scoring lowest on this. I'll leave this out of the discussion for now.
The more criteria are required, the higher the specificity of the diagnosis. In other words, one will be surer that the person diagnosed will in fact have leprosy. Therefore, one would be most sure (highest specificity) if all 3 cardinal signs were present. But because sensitivity and specificity are 'in balance', it will also mean a lower sensitivity and therefore a proportion of people who in fact have leprosy will be misdiagnosed as not having leprosy, because they only have one or two signs. The same argument holds when comparing two cardinal signs versus one. Requiring two signs for a definitive diagnosis, as suggested by Dr Naafs, will result in fewer false positive diagnoses, but more false negative diagnoses ('missed cases'), when compared to requiring only one sign to be present.
In addition, the predictive value of the diagnosis is considered important (particularly the positive predictive value - PPV). This represents the likelihood that someone with the given cardinal sign(s) does in fact have leprosy. The PPV is dependent on the prevalence of the condition. For a given sensitivity and specificity combination, the PPV will be higher with a higher prevalence in the population examined. For example, the PPV of a one-sign criterion (only one cardinal sign is sufficient to diagnose leprosy) will be high in a group of leprosy suspects or contacts of known leprosy cases. But it will be much lower in a sample of the general population. Likewise, it will be higher in a leprosy-endemic country than, e.g. in the Netherlands. This is why the consensus to date has been that, in leprosy-endemic countries, the presence of one cardinal sign is sufficient to diagnose a case of leprosy.
Given the above, it could be argued that this criterion is too simplistic, since there are large differences in prevalence/incidence between different endemic countries. In addition, there are large differences in prevalence of some of the differential diagnoses mentioned by Dr Naafs. The likelihood of someone in Indonesia with a positive skin smear having Buruli ulcer is negligibly small, while in Benin, such a person may be more likely to have Buruli than leprosy.
For field purposes, however, the requirement of two cardinal signs being positive is not practical, since most health workers do not learn to palpate nerves and skin smear services are not available or only to a selected few suspects who need special referral to have such a test done. While referral centres may have several options at their disposal to (further) improve the certainty of a diagnosis of leprosy, I propose that for integrated control programmes, one 'cardinal' sign should remain sufficient. This would cause us to err on the safe side, with a relatively high sensitivity and low specificity (few missed diagnoses). Under very low-endemic circumstances, or high prevalence of a condition easily mistaken for leprosy, additional criteria will be needed. In addition, if the risk of stigmatisation and social exclusion based on the diagnosis of leprosy is very high, then one may want to require additional assurance before labelling someone as having leprosy. This could be in form of a second opinion, rather than additional tests.
With best wishes,
Wim van Brakel
NLR Technical Advisor,
Leprosy Unit
Royal Tropical Institute
Amsterdam
LML - S Deepak, B Naafs, S Noto, P A M Schreuder
LML Archives: http://www.aifo.it/english/resources/online/lml-archives/index.htm
Dr Salvatore Noto
Padiglione Dermatologia Sociale
Ospedale San Martino
Largo R. Benzi, 10
16132 Genoa, Italy
Tel: (+39) 010 555 27 83 - Fax: (+39) 010 555 66 41 - E-mail: salvatore.noto@hsanmartino.it
No comments:
Post a Comment