P1.15-06 Integrative Quality Improvement for Management of Lung Cancer in Uganda

Abstract

V769_D770InsASV, H773_V774InsH, H773_V774insNPH, and A763_Y763insFQEA. Conclusion: The frequency of EGFR exon 20 insertions was 0.1À2.1% of NSCLC patients, with a high variability in both length and position of insertions within exon 20. Currently, available data are sparse and come primarily from studies based on single-center experiences, with data gaps across several large geographic regions and populations. Results also indicate a need to further explore underlying geographic variations in epidemiology. Larger, multi-center global studies will further help to refine the frequency of exon 20 insertions and other uncommon mutations in NSCLC.Background: Uganda is currently engaged in quality improvement initiatives for cancer patients. One of them is the creation of an integrative quality system, consisting of guideline development, quality indicators definition and feedback to hospitals. This approach has already been successfully implemented for five types of cancers: rectum (in collaboration with development partners), breast, testis, oesophagus and stomach. Building on previous experience, the study presents the development of a set of quality indicators (QIs) for the management of lung cancer. Method: We followed the standardized MOH methodology to identify, select, test and measure the indicators. Because patients may be in contact with different hospitals (for instance, be diagnosed in one hospital but receive treatment e surgery or radiotherapy e in another), we developed a specific algorithm to attribute each patient to the centre where he/she was diagnosed or received treatment (surgical centre or centre of radiotherapy). The method to test the feasibility of identifying comorbidities of patients based on their pharmaceutical billing data during the year before the cancer diagnosis is also described. Result: The results of this project made clear that we need more complete and accurate reporting of data to allow more precise and correct evaluation of the quality of care for lung cancer patients in Uganda. Quick and fluent data collection would make it possible to provide comprehensive feedback to care providers on a regular and timely basis. To make this happen, investments in data registration and analysis will be necessary. Conclusion: When benchmarking results between hospitals, cautious interpretation is warranted. The use of funnel plots avoids spurious ranking of hospitals and outlier dots can reliably designate either good or bad performers. Statistical modeling can often only partially account for differences in case-mix and other biases. Judging quality of care delivered by a hospital is further hindered by the often small number of patients treated per hospital. Hence, from a sheer statistical point of view, small volumes of activity make it impossible to offer an acceptable level of assurance about the quality delivered to the patient.Background: Special attention is required by elderly population, due to chemotherapy (ch) risks and comorbidities, which may limit the capacity to deliver opti...