G458(P) Coding of admission diagnosis in a district general hospital in myanmar

Background With increasing numbers of children presenting to state and district general hospitals, including those transferred from smaller township and station hospitals, there is increasing need to understand the burden of disease presenting to hospital. The WHO recommends all countries use ICD-10 coding for reporting health data. There is little research to date describing the terms used to record admission diagnoses in paediatrics in Myanmar. Aim To review the library of terms currently used to record admission diagnosis in a typical paediatric ward and map these against the ICD-10; to assess feasibility of applying this standardised system of coding across paediatrics in Myanmar. Methods Logbook records from the Child Ward (CW) were reviewed for the period May 2018 – April 2019. All primary admission diagnoses were recorded and grouped by body system. Without editing or inferring meanings of the terms, these were individually tested for matching against ICD-10 codes. Results Over a 12 month period, 343 different admission diagnoses were recorded. It was possible to match 90.9% diagnoses to ICD-10 codes. However, many could only be matched to ‘not specified’ codes and 18.5% could only be matched to terms in Chapter XVIII (symptoms and signs not elsewhere classified). Some chapters had no matches across the 1 year review period. Conclusions We have shown that it is possible to use ICD-10 but the application of such a complex system has multiple challenges in a resource-limited setting. The heavy use of ‘not specified’ codes reflects a lack of diagnostic precision, which is multifactorial, influenced by factors including limited availability of investigations, variable access to specialist advice and lack of training in coding. A coding system for this setting would need to factor in the inherent diagnostic uncertainty and be user-friendly enough to require little clinician training, whilst yielding clinically and academically robust data. We propose that a standardised approach to diagnosis recording, with modifications to overcome the challenges of directly applying ICD-10 in a low-resource setting, would be useful here and in other global healthcare facilities to guide resource allocation and to monitor changes in patterns of illness according to interventions.