Fluid management knowledge in hospital physicians: ‘Greenshoots’ of improvement but still a cause for concern

Discussion

This study suggests that the fluid management ‘knowledge base’ is poor across the whole ‘medical’ workforce including middle grade and senior physicians. Essential, desirable and overall knowledge scores were significantly lower for all grades compared to ICCs (the best available comparator group). Despite poor overall scores, with a median of only two of five ‘essential’ and two of five ‘desirable’ knowledge questions answered correctly, there were ‘greenshoots’ of improvement compared to previous assessments.

In postoperative patients, poor fluid and electrolyte management prolongs hospital stay and increases cost, morbidity and mortality.^{3–5,13–16} Between 17–54% of postoperative cases experience at least one fluid related complication due to under/over hydration and/or harmful electrolyte imbalance (eg tachyarrhythmia).^3,14,15 The poor fluid management knowledge of junior clinicians is well documented and has remained a concern in recent studies.^6,15–19 Surgical SpRs and consultants perform slightly better than their juniors but little is known about the knowledge of ‘senior medical’ clinicians.²⁰ In postoperative surgical patients, junior clinicians are responsible for >85% of fluid prescriptions.^1,3,20 However, in medical practice, middle grade clinicians (eg SpRs) often ‘direct’ fluid management, which is subsequently ‘written up’ by junior colleagues. In an ‘apprenticeship-style’ training system, like GIM, inadequate senior clinician ‘knowledge’ is likely to impact junior colleagues training.

Our study is concerning but provides some optimism, too. FY and CMT fluid management ‘knowledge’ could be compared with results from four previous studies between 2001–2016.^1,6,11,12 Identical or very similar questions were used (and the data recorded) in these previous reports and our study. For example, 65% FY and 73% CMT correctly identified the sodium chloride content of 0.9% saline in our study compared to 21–37% and 46–51%, respectively, in previous reports. Reassuringly, this question was also correctly answered by >75% of SpRs and consultants. In addition, ∼50% of FYs in this study knew the recommended daily water and sodium requirements compared to 6–33% in previous reports. Also, 34% of FYs correctly identified the composition of Hartmann's solution compared to 8–13% previously, although less improvement was seen for Gelofusine composition (9%; previously 4%), probably related to recent reductions in the use of colloid solutions. Earlier this year, a similar study in FYs also reported an improvement in the knowledge of daily water/electrolyte requirements but not of fluid composition (only 31% knew the sodium content of 0.9% saline).²⁰ ‘Desirable’ knowledge also improved compared to historical assessments, with most clinicians aware of the metabolic risks of excess saline. However, less than 50% knew the normal serum chloride or intracellular electrolyte concentrations. Overall ICU SpRs scored better than GMCs and medical SpRs, possibly due to training factors or greater ICC knowledge in an ‘apprenticeship teaching system’. Nevertheless, despite the improvements reported above, 50% or more of medical practitioners still did not know the daily fluid and electrolyte requirements recommended in NICE CG174 or the electrolyte compositions of commonly used fluids (supplementary material S3).

In four questions using NICE CG174 recommendations to test ‘typical clinical fluid management scenarios’, the best intravenous fluid and electrolyte ‘maintenance’ regimen for an elderly comorbid patient unable to drink was correctly identified by 75% of ICCs but, disappointingly, only 17–30% of ‘medical’ practitioners. However, two questions addressing the best ‘replacement’ regimen in a complex postoperative patient with large nasogastric losses and the recommended use of a physiologically balanced fluid for a high-volume resuscitation were correctly answered by most senior/middle grade medical practitioners (but <50% of FYs). In the final complex question, ≤50% of respondents recognised that the oliguria in a recovering trauma case was due to reduced intravascular volume.

Junior clinicians’ fluid management confidence has previously been reported to be greater than ‘knowledge test’ results suggest is reasonable.^1,6,12 In this study, pre-test confidence was weakly but significantly correlated with test results across all participants (Table 2). This correlation was stronger in ICCs (but not significant). When ICC participants were excluded from the analysis, the correlation in remaining participants was weaker, but still significant, consistent with some pre-test recognition of knowledge ‘weakness’. In terms of patient safety, lack of awareness of knowledge deficits, described as a failure of metacognition, is a concern. Metacognition describes the awareness of one's own knowledge and the ability to reflect on, evaluate and address deficits in self-knowledge.^21,22 Although less of an issue in the specialty aspects of a physician's practice, it probably does impact on ‘general’ patient care (eg nutrition or falls management).²¹ Failure of metacognition is as likely to occur in senior clinicians as it is in junior clinicians.

SAEs reviews effectively address ‘knowledge deficits’, improves practice, prevents recurrence and enhances team learning. This study demonstrates a failure of this ‘safety’ mechanism in fluid management. Although most practitioners frequently manage fluid-related SAEs, and ∼50% of senior clinicians have caused a SAE, few were reported or investigated. Consequently, this vital aspect of learning is lost. Improving SAE review presents an opportunity to promote learning and partly addresses the failure of the ‘apprenticeship system’. The ‘test, teach and retest’ strategy was also an effective training tool as demonstrated in the 40 participants who redid the test, answering an additional nine questions correctly.

There were study limitations. Although trainee and ICC numbers were robust, it was difficult to ‘persuade’ GMCs to participate. Only 11 GMCs completed the questionnaire (31%). This limited consultant participation and variability in group sizes is a weakness. In addition, few senior colleagues (∼20%) were prepared to complete the questionnaire supervised. Other issues included the reduction in ‘colloid fluid’ use during the study, rendering an ‘essential’ knowledge question about Gelofusine less relevant and, for reasons of anonymity, post teaching scores could not be matched. Finally, data collection occurred over a 3-year period, potentially allowing increased ‘penetrance’ of NICE guideline knowledge into clinical practice. Although this appeared to be a factor, we were unable to demonstrate a significant effect.

This study suggests that fluid management knowledge remains inadequate across the hospital medical workforce, including senior colleagues. However, improvement has occurred in specific areas (eg daily water/electrolyte requirements) compared to historical reports. The weak correlation between confidence and ‘test’ scores suggests physicians do have awareness of this issue, but in these circumstances traditional ‘apprenticeship systems’ of learning are unlikely to be effective, compounded by poor training and failure to learn from SAEs. We propose that undergraduate and postgraduate training should be standardised and based on NICE CG174. It should be delivered by those with appropriate knowledge and adequately assessed within training programmes. Senior clinicians need to be made aware of this potential deficiency in their knowledge base and appropriate training material made available with greater emphasis on this topic in ‘refresher’ courses and ‘specialist’ conferences.