Editorial: Keeping it simple: the power of three clinical protocols

Recently, colleagues and I published in the Lancet, results from a cluster randomised controlled trial that provided high-level evidence that good nursing care makes a difference to patient outcomes (Middleton et al. 2011b). This is one of the few trials internationally ever to demonstrate that nursing care can positively influence the ‘hard’ clinical endpoints of death and dependency and as such has made a noteworthy contribution to research in the care of the acute stroke patients. Our trial was an example of implementation research aimed to change clinician behaviour for the management of fever, hyperglycaemia and swallowing difficulties and was tested in the clinical area of acute stroke, yet, our results have implications for clinical practice beyond stroke. As outlined in our paper (Middleton et al. 2011b), we developed and implemented three clinical protocols for fever, hyperglycaemia and swallowing management (known as the Fever, Sugar Swallowing or FeSS protocols). Each protocol was developed by multidisciplinary panels of experts and were based on evidence and guided by recommendations in our national stroke clinical practice guidelines (National Stroke Foundation 2010), all pertained to the first 72 hours of stroke unit care. The fever protocol asked nurses to monitor patients, temperature four hourly and to treat episodes of fever (temperature > 37·5 °C) with paracetamol. The hyperglycaemia protocol included a formal venous blood glucose level on admission; 1–6-hourly monitoring of blood glucose levels (frequency varied by glucose reading with more frequent readings required where glucose was elevated); treatment of initial hyperglycaemia (8–11 mmol/l for known diabetics; 8–16 mmol/l for non-diabetics) with saline for six hours; and at anytime thereafter insulin administration for hyperglycaemia (≥11 mmol/l for known diabetics; ≥16 mmol/l for non-diabetics). Our swallowing protocol consisted of nurses undergoing an education programme to be deemed ‘swallow screeners’. This educational programme was run by speech pathologists at each individual intervention site who were provided with a standardised learning package consisting of a DVD, power point presentation, knowledge test and a competency test. Nurses had to accurately undertake three swallow screens as witnessed by a speech pathologist. Once nurses had undergone this training, they were asked to screen all patients for swallowing difficulties within the first 24 hours of admission to the stroke unit and refer those who failed the screen to a Speech Pathologist for a swallowing assessment. However, the mere existence of these protocols does not guarantee their successful uptake into routine clinical care (Grimshaw et al. 2004). Hence, we designed an implementation strategy (FeSS Implementation Strategy) to foster multidisciplinary teamwork and adoption of the FeSS protocols. Our FeSS Implementation Strategy consisted of two multidisciplinary team-building workshops, working with clinicians from intervention stroke units to overcome barriers to the implementation of the FeSS protocols ranging from lack of thermometers and syringe pumps, to concern over increased workload. Next, we ran two site-based educational sessions and supported local stroke unit co-ordinators with site visits, telephone and email reminders. Control group stroke units received only an abridged version of the national stroke guidelines relevant to fever, hyperglycaemia and swallowing management. Nineteen acute stroke units in New South Wales, Australia and a total of 1696 patients participated in the trial. Ten stroke units were randomised to the intervention group and nine to the control group. We rigorously evaluated the effect of our intervention using a cluster randomised controlled trial design. We allowed three months for our intervention to become ‘bedded down’ as part of ‘usual care’ in the relevant stroke units and then evaluated patient outcomes and processes of care. Our results showed that patients who received care in acute stroke units randomised to receive our intervention were 15·7% more likely to be alive and independent 90 days following their stroke (modified Rankin Score ≥2) and had improved physical functioning (SF-36 mean physical component summary score), requiring only six patients to be treated to prevent one patient from dying or becoming dependent (number needed to treat [NNT] = 6·4) (Middleton et al. 2011b). These results were remarkable and exceed the published benefits of all of the three current proven treatments for stroke, namely aspirin (NNT = 79; Sandercock et al. 2008), stroke unit admission [NNT = 18; Stroke Unit Trialists’ Collaboration 2007) and thrombolysis [(NNT = 8; The National Institute of Neurological Disorders & Stroke rt-PA Stroke Study Group 1995); (NNT = 14; Hacke et al. 2008), depending on the time of stroke onset-to-treatment]. Further, patients who received care in our intervention stroke units had lower mean temperatures, fewer episodes of fever, lower mean glucose readings and improved swallowing screening rates. Excitingly, our results have generated great interest in the international literature including a fast-tracked publication by the Lancet with an accompanying commentary (‘an innovative cluster trial…; Importantly, the trial … also uses implementation science theory to maximise the chances of the intervention being sustainable in the longer term’; Wolfe & Rudd 2011) a commentary published in the Annals of Internal Medicine; ‘well done study’; ‘very positive results’; Alberts 2012), inclusion in the library of the Faculty of 1000 following postpublication peer review placing the work in the top 2% of international articles in biology and medical research (‘unique design…well conducted…findings should be taken very seriously’; Norrving 2011). The trial also won the 2011 Canadian Stroke Congress Award for Impact. In addition, commentaries by our team have been published targeting various clinical interest groups (Middleton 2012a and b). All this coverage demonstrates the keen interest and relevance of our study results amongst different clinical groups as our message is internationally relevant and not just limited to the area of stroke. So what are the implications of our results for clinical practice? Of importance was the pragmatic nature of the FeSS clinical protocols. They were not complicated and guided nurses in three practice areas that are paramount for good nursing care centred around managing two important complications (fever and hyperglycaemia) and keeping patient's safe (swallowing management). We did not aim to tightly control glucose levels unlike other stroke studies (Gray et al. 2007) and, indeed, when we commenced our intervention education, many nurses told us they already were providing care similar to that proposed in our protocols. However, our pre-intervention audit showed this not to be the case (Drury et al. 2010) outlining the importance of continual monitoring to ensure we really are delivering care based on best practice at all times. As stated in the Lancet commentary by Wolfe and Rudd, ‘if the Australian findings are transferable to the northern hemisphere, then we would hope to see dramatic effects across whole cities’ (Wolfe & Rudd 2011). In my role as a clinical chair, I am often asked by nurses ‘Should I change my care based on the results from one randomised controlled trial?’ The answer may frequently be ‘No’ because of the difficulty generalising the findings. For example, the sample size may be small or the study excluded significant groups of the population or has other evidence of bias. In our study, the sample size was large and our pragmatic intervention was evidence-based with impressive positive effects for patients. Thus, in this instance, patients would be well served if nurses integrated our findings into their every day clinical practice. In addition, whilst tested in the gold standard setting of an acute stroke unit, they also may be useful to guide clinical practice in smaller hospitals who do not have a stroke unit. Evidence-based fever, hyperglycaemia and swallowing management also are not issues specific only to stroke, and hence, the protocols themselves could have wider applica-bility beyond the clinical specialty of stroke. Certainly, this is the case for our implementation strategy. Changing clinician behaviour with a multi-faceted intervention like ours using barrier identification (Grol et al. 2005), reinforcement of multidisciplinary teamwork (Hamilton et al. 2007), local adaptation (Grol & Grimshaw 1999) and use of site champions (Flodgren et al. 2007) is highly transferable and would be worthy of replicating to test implementation of similar nurse-initiated clinical protocols in any clinical setting. Nursing as a profession is continually evolving. We have increasing specialisation and new clinical practice roles that often involve an increased scope of practice such as prescribing medications and ordering diagnostic tests. The future for nursing is exciting and the possibilities not yet fully realised in terms of new options for how to provide better care for patients. However, our trial has shown that along the way we must not forget the simple things. Unless all nurses are consistently providing this care to all patients on every shift, we will not see a population benefit for our patients (Middleton 2012a). Successful implementation of these three protocols premised on evidence-based nursing care had a highly positive effect on patient outcomes and, as stated in the Faculty of 1000 postpeer review publication, ‘careful attention to aspects of nursing and general care appears to generate large benefits’ (Norrving 2011). As a profession, we must seize opportunities for expansion and change in our role but not lose sight of the value of doing the simple things well, that is timely patient observations and provision of prompt, evidence-based nursing care. Our trial has shown that such care can save lives.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.