Making Clinical Practice Guidelines Pragmatic: How Big Data and Real World Evidence Can Close the Gap

Clinical practice guidelines (CPGs) have become ubiquitous in every field of medicine today but there has been limited success in implementation and improvement in health outcomes. Guidelines are largely based on the results of traditional randomised controlled trials (RCTs) which adopt a highly selective process to maximise the intervention's chance of demonstrating efficacy thus having high internal validity but lacking external validity. Therefore, guidelines based on these RCTs often suffer from a gap between trial efficacy and real world effectiveness and is one of the common reasons contributing to poor guideline adherence by physicians. "Real World Evidence" (RWE) can complement RCTs in CPG development. RWE-in the form of data from integrated electronic health records-represents the vast and varied collective experience of frontline doctors and patients. RWE has the potential to fill the gap in current guidelines by balancing information about whether a test or treatment works (efficacy) with data on how it works in real world practice (effectiveness). RWE can also advance the agenda of precision medicine in everyday practice by engaging frontline stakeholders in pragmatic biomarker studies. This will enable guideline developers to more precisely determine not only whether a clinical test or treatment is recommended, but for whom and when. Singapore is well positioned to ride the big data and RWE wave as we have the advantages of high digital interconnectivity, an integrated National Electronic Health Record (NEHR), and governmental support in the form of the Smart Nation initiative.

Use of endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) in the diagnosis of granulomatous mediastinal lymphadenopathy

<p><b>INTRODUCTION</b>This study assessed the clinical utility of endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) for the diagnosis of suspected granulomatous mediastinal lymphadenopathy.</p><p><b>MATERIALS AND METHODS</b>Retrospective chart review of all patients who underwent EBUS-TBNA for suspected granulomatous mediastinal lymphadenopathy at Singapore General Hospital between December 2008 and December 2011 inclusive.</p><p><b>RESULTS</b>Over a period of 3 years, a total of 371 patients underwent EBUS-TBNA of whom 33 (9%) had the procedure performed for evaluation of suspected granulomatous mediastinal lymphadenopathy - 18 for suspected tuberculosis (TB) and non-tuberculous mycobacterial (NTM) lymphadenitis, and 15 for suspected sarcoidosis. EBUS-TBNA was diagnostic in 9 of the 13 patients with a final diagnosis of TB/NTM. EBUS-TBNA cultures were positive in 6 of them (46%), 1 showed acid-fast bacilli (AFB) although cultures were negative, and 2 had necrotising granulomatous inflammation from EBUS-TBNA biopsies and sputum cultures grew TB. EBUS-TBNA was diagnostic in 9 of the 14 patients with a final diagnosis of sarcoidosis through histology showing non-caseating granulomatous inflammation. The sensitivities of EBUS-TBNA for diagnosis of TB/NTM, sarcoidosis and overall granulomatous mediastinal lymphadenopathy were 69%, 64%, 64%; the negative predictive values were 56%, 17%, 33%; and accuracies were 78%, 67%, 70%, respectively.</p><p><b>CONCLUSION</b>EBUS-TBNA can be useful in the diagnosis of suspected granulomatous mediastinal lymphadenopathy with sensitivities and accuracies of >60%.</p>