METHODS: This prospective cohort study was conducted at Hospital Universiti Sains Malaysia (USM) from 1st October 2016 to 30th November 2017. Serial bedside ultrasound procedures were performed for 83 patients who were diagnosed as having dengue fever without warning signs and were initially treated as outpatients. Ultrasonography evidence of plasma leakage either pleural effusion, thickened gallbladder wall, ascites or pericardial effusion were compared with clinical findings and laboratory parameters for plasma leakage.
RESULTS: Of the 83 dengue patients, eventually 72.3% had dengue fever with warning signs and 6.0% had severe dengue fever. There were 38 patients who had subclinical plasma leakage at initial presentation, 84.2% and 7.9% of them then progressed to dengue fever with warning signs and severe dengue respectively. There was a minimal agreement between serial bedside ultrasound and haematocrit level in the detection of plasma leakage (observed kappa 0.135).
CONCLUSIONS: Serial bedside ultrasound is an adjunct procedure to physical examination and may detect plasma leakage earlier compared to haemoconcentration. The early usage of serial ultrasound is of paramount importance in detecting dengue patients who are at risk of progressing to severe dengue.
MATERIALS AND METHODS: The panel consists of ten experts from ED that was randomly chosen from list of specialists obtained from the National Specialist Registry for Emergency Medicine. A set of questionnaires that contains item constructs related to KPI based on structure, outcome and process was developed from initial literature search from Pubmed Central, Google Scholar, Cochrane Database and Public Library of Sciences. The construct then used for FDM session in second phase of the study. In FDM phase, the experts will rank each of the items created from nominal group technique (NGT) session by using Likert Scale ranged from 1 to 5 ("1" totally disagree and "5" extremely agree). FDM prerequisite must include threshold value (d) ≤0.2, expert consensus of >75% and average fuzzy numbers ("A" value) of >0.5.
RESULTS: The initial item construct has produced 22 items proposed for the service KPI. Post FDM analysis for service KPI, 16 out of the 22 (72%) satisfied first prerequisite "d" value ≤0.2. For the second prerequisite, ten items (45%) from service KPI domain had expert consensus of more than 75%. For the third prerequisite, 16 out of the 22 (73%) fit the criteria of average fuzzy number ("A" value) of more than 0.5. In final model of FDM, 13 items (59%) were discarded and the remaining (n=9 items) that fulfilled all three prerequisites were retained for the final draft for content validation process.
CONCLUSION: This study introduces that FDM can be used to obtain experts' opinion and consensus in order to achieve a decision. The experts' consensus on the suitability of the pre-selected items on the KPI set were obtained, hence it is now ready for further applicability in the clinical setting in ED.