Improving clinical governance of HIV treatment programmes in resource poor settings: the role of digitising clinical notes

High standards of clinical governance are crucial for providing sustained and effective health care in resource poor settings. However, implementing clinical audit in these settings is often beset by practical issues such as staff shortages and inadequate support or supervision, and is often seen as burdensome, rather than being a tool for improvement in practice.^1,2 Out-sourcing the burden of data analysis may go a long way to facilitating regular audit in a resource poor setting. Critical to this is the development of a cheap and quick system to gather data for remote auditing.

We investigated the feasibility of using an inexpensive 12-megapixel point-and-shoot digital camera to collect data from clinical notes in a format capable of being sent via secure electronic file transfer for remote analysis. We photographed a random selection of 20 anonymised clinical records, from among 1000, of the HIV outpatient clinic of Cecilia Makiwane Hospital (CMH), East London, South Africa – the main hospital in the township of Mdantsane, with 750 beds, 80 doctors and serving a population of one million. We measured the time taken to retrieve and photograph each record then evaluated the digital images for legibility and whether information on use of the antiretroviral drug tenofovir, suitable for clinical audit, could be retrieved from these images at a later time. Specifically, we audited whether antiretroviral drug name and creatinine clearance were recorded at baseline, months 3 and 6, and then every 12 months thereafter, in line with the latest South African Department of Health Standard Treatment Guidelines. The South African Department of Health guidelines ³ recommend against the use of tenofovir in patients with a creatinine clearance less than 50 ml/min. Prolonged use of this agent can lead to reduced renal function. ⁴ Monitoring creatinine clearance allows clinicians to switch timeously to alternative agents.

This work was approved by the East London Hospital Complex Ethics Committee.

The time to retrieve and photograph each of the 20 records ranged from 90 s to 5 min, with a mean of 2.5 min. Only one out of the 20 records photographed was unreadable due to operator error. Seven of the patients were receiving tenofovir. None of the legible records indicated, either in the notes or by inference from results, that blood for creatinine clearance measurement had been taken according to clinical guidelines.

Major global efforts have been made in recent years to make ART widely available to populations in resource poor settings ⁵ with evidence-based treatment guidelines leaning increasingly towards treating earlier in the course of infection. These advances have major implications for drug supply and budgets ⁶ particularly in a restricted economic climate. ⁷ In this proof of concept feasibility study we demonstrated a reliable, cheap and quick method to digitise clinical records in an electronically transferable format, offering a way to facilitate standardised and large-scale collection of auditable data. These data files could be transferred over the internet or a mobile phone network, if a high bandwidth connection is available. They could also be saved to a CD, DVD or a secure memory stick – all of which are comparatively low-cost options available in even the poorest of sub-Saharan hospitals. We demonstrated that these images could be analysed remotely, and propose that this could be done cheaply, and on a routine basis, either by groups of trained volunteers (such as retired, or overseas clinicians and others experienced in clinical audit), or by in-house staff. Furthermore, we showed that the images are of high enough quality to perform a pre-defined simple clinical audit. Our study has stimulated further research in this area at CMH.

This approach is a flexible and affordable way to collect clinical data for audit in any low-income setting. Key challenges include: (1) building relationships with, and maintaining, a reliable bank of auditors to analyse records in a timely fashion; (2) appointing clinical audit staff with specific responsibility for collecting images and forwarding them to auditors for analysis, and (3) ensuring audit results are received promptly and fed back to practitioners to complete the cycle of quality improvement.

More work is required to assess whether these methods can improve clinical practice and whether they are applicable across auditable outcomes both for HIV and other chronic diseases in resource poor settings.