Taal Anna Theodora, Post Erik B, Hussaini Tijjani, Barminus Augustin Gayus, Dahiru Tahir
Netherlands Leprosy Relief, Amsterdam, Netherlands.
The Royal Tropical Institute, Amsterdam, Netherlands.
Res Rep Trop Med. 2015 Oct 2;6:85-94. doi: 10.2147/RRTM.S87600. eCollection 2015.
Skin diseases are common worldwide, though prevalence rates in rural areas are difficult to estimate, and are primarily based on hospital studies rather than community-based studies. Primary health care providers in rural areas often lack sufficient knowledge about skin diseases, which contributes to poor skin management and subsequently causes considerable morbidity. This study looked at the performance of first-line health care providers in the management of common skin disease, using an algorithmic approach with a flowchart with diagnostic steps. As a reference standard, two dermatologists independently validated the diagnoses and treatment choices made by the providers. The performance of the algorithm was calculated in terms of the sensitivity, specificity, positive predictive value (PPV), and negative predictive value for each skin disease of the algorithm. A total of 19 patent medicine vendors and 12 traditional healers from Kano State in Nigeria diagnosed 4,147 patients with suspected skin symptoms. The most common skin disease was tinea capitis (59.2%), and it was found predominantly among boys below 15 years of age. Together, patent medicine vendors and traditional healers had 82% of the cases correctly diagnosed, and in 82% they prescribed the correct treatment. The sensitivities varied for each skin disease from 94.8% for tinea capitis to 7.1% for contact dermatitis. The specificities varied between 87.0% and 98.6%. Except for tinea capitis, lower PPVs were found for the various skin diseases when compared to earlier studies. In spite of the observed low sensitivities and low PPVs for several diseases, the algorithm seems to offer an improvement in management of common skin diseases at the peripheral level. With adaptations in training, further refinement of the algorithm and refresher training, predictive values and sensitivities can be increased.
皮肤病在全球都很常见，不过农村地区的患病率难以估计，主要是基于医院研究而非社区研究。农村地区的初级卫生保健提供者往往缺乏足够的皮肤病知识，这导致皮肤管理不善，进而造成相当高的发病率。本研究采用带有诊断步骤流程图的算法方法，观察一线卫生保健提供者对常见皮肤病的管理表现。作为参考标准，两名皮肤科医生独立验证了提供者做出的诊断和治疗选择。根据算法对每种皮肤病的敏感性、特异性、阳性预测值（PPV）和阴性预测值来计算算法的性能。来自尼日利亚卡诺州的19名成药销售商和12名传统治疗师诊断了4147名有疑似皮肤症状的患者。最常见的皮肤病是头癣（59.2%），主要发现在15岁以下的男孩中。成药销售商和传统治疗师合计正确诊断了82%的病例，并且在82%的病例中开出了正确的治疗方案。每种皮肤病的敏感性各不相同，头癣为94.8%，接触性皮炎为7.1%。特异性在87.0%至98.6%之间。与早期研究相比，除头癣外，各种皮肤病的阳性预测值较低。尽管观察到几种疾病的敏感性和阳性预测值较低，但该算法似乎在外围层面改善了常见皮肤病的管理。通过调整培训、进一步完善算法和进行复习培训，可以提高预测值和敏感性。
