- Email: [email protected]
Dysphagia in acute stroke: a long-awaited trial
Reflection and Reaction
Dysphagia in acute stroke: a long-awaited trial See Articles page 31
16
Carnaby and colleagues1 describe in this issue of The Lancet Neurology the results of their single centre, singleblind, randomised controlled trial of a standard behavioural intervention for dysphagia after stroke as “encouraging”. I might even add the adjective exciting, because this trial has shown, for the first time, that an intervention can improve outcomes of patients with dysphagia. Dysphagia is more common in severe stroke, but is independently associated with poorer long-term outcomes in mild-to-moderate stroke.2 This trial indicates that a behavioural intervention not only improves recovery of swallowing, but also reduces the complications of dysphagia and improves long-term outcomes. This finding lends support to the notion that the association between dysphagia and poor outcomes is, at least in part, causal. Because almost half of all patients admitted to hospital with an acute stroke worldwide have dysphagia an effective treatment would be applicable to millions of patients each year, and therefore might have huge public-health benefits. However, before we get overly excited we must consider the limitations of this study, some of which the authors describe in their discussion. There are several examples, even in the stroke specialty, of small, single-centre, randomised trials reporting positive results that are then not confirmed by larger multicentre trials. Two examples are the effectiveness of percutaneous endoscopic gastrostomy feeding after acute stroke3,4 and leisure therapy.5,6 Many small trials that yield overly optimistic results have methodological failures that can introduce bias. The most common problems are: failure to enrol enough patients to ensure that randomisation produces treatment groups with similar baseline characteristics; failure to use secure methods of randomisation that prevent any foreknowledge of treatment allocation; failure to properly blind outcome assessments; failure to follow up all patients and to include all patients randomised in the primary analyses; failure to stipulate in advance what the primary outcome is so that the analysis becomes a data dredging operation, which is likely to produce false positive results; and failure to prespecify if any, or how many, interim analyses will be undertaken so that the trial is stopped simply when the chief investigator identifies a significant effect that they
think will make their trial acceptable to the editors of journals. Carnaby and colleagues have avoided most, if not all, of these failures. They used a secure randomisation system to enrol a respectable number of patients who were well described at baseline. They prespecified their sample size, which they achieved, and prespecified their primary and secondary outcomes, which were assessed largely blind to treatment allocation. They were rigorous in follow-up and accounted for all patients in their analysis. Moreover, they should be congratulated for describing their trial, their interventions, and their results in a way that would allow others to replicate their work. We now need more precise estimates of the effectiveness of the interventions and whether more intensive, and thus presumably more expensive, behavioural interventions are more effective than less intensive ones. We also need to determine whether this behavioural intervention, delivered by less specialist, skilful, or enthusiastic speech and language therapists working in many different environments, can achieve similar improvements in clinically relevant outcomes. Determining whether videofluoroscopy is an essential component or whether it could be dispensed with or replaced by other techniques, such as fibreoptic endoscopic assessment of swallowing, would also be useful. To reliably answer these questions will need many hundreds, or more likely thousands, of patients to be recruited into randomised trials. These patients might be enrolled in further single-centre trials, with similar methods to those described by Carnaby and colleagues, which could then be summarised in an informative meta-analysis. However, a family of large multicentre trials including a factorial design to address the questions of videofluoroscopy or alternatives would be a better way forward. I am therefore frustrated and somewhat disappointed that this trial has taken almost 5 years to report its results. These are 5 wasted years when further studies could have been planned, funding identified, and recruitment started. If we are to deliver better treatment to our patients we have to streamline our research efforts, avoiding such delays in reporting encouraging results and establishing confirmatory http://neurology.thelancet.com Vol 5 January 2006
Reflection and Reaction
studies. Hopefully, initiatives such as the Stroke Research Networks, funded through the UK Clinical Research Collaboration and Chief Scientist Office in Scotland, which aim to provide infrastructure to facilitate multicentre stroke trials, will help us to assess such promising interventions more quickly. Delays in introducing thrombolytic therapy as a standard treatment for myocardial infarction worldwide, which arose from failures to summarise existing data indicating its effectiveness and to carry out really definitive trials earlier, have been estimated to have killed thousands of patients.7 I hope that our delay in confirming the effectiveness of a standard behavioural intervention for stroke patients with dysphagia does not prove to be as costly. Martin Dennis
Division of Clinical Neurosciences, University of Edinburgh, Western General Hospital, Edinburgh EH4 2XU, UK [email protected]
See UK Clinical Research Collaboration home page: http://www.ukcrc.org/
I have no conflicts of interest. 1 2 3
4
5 6
7
Carnaby G, Hankey GJ, Pizzi J. Behavioural intervention for dysphagia in acute stroke: a randomised controlled trial. Lancet Neurol 2005; 5: 31–37. Smithard DG, O’Neill PA, Park C, et al. Complications and outcome after acute stroke: does dysphagia matter? Stroke 1996; 27: 1200–04. Norton B, Homer-Ward M, Donnelly MT, Long RG, Holmes GKT. A randomised prospective comparison of percutaneous endoscopic gastrostomy and nasogastric tube feeding after acute dysphagic stroke. BMJ 1996; 312: 13–16. FOOD Trial Collaboration. Effect of timing and method of enteral tube feeding for dysphagic stroke patients (FOOD): a multicentre randomised controlled trial. Lancet 2005; 365: 764–72. Drummond A, Walker MF. A randomised controlled trial of leisure therapy after stroke. Clin Rehab 1994; 9: 283–90. Parker CJ, Gladman JR, Drummond AE, et al. A multicentre randomized controlled trial of leisure therapy and conventional occupational therapy after stroke: Trial of Occupational Therapy and Leisure. Clin Rehab 2001; 15: 42–52. Lau J, Antman EM, Jimenez-Silva J, Kupelnick B, Mosteller F, Chalmers TC. Cumulative meta-analysis of therapeutic trials for myocardial infarction. New Engl J Med1992; 327: 248–54.
Phenobarbital for convulsive epilepsy at primary care level
http://neurology.thelancet.com Vol 5 January 2006
had one of all adverse effectss between 19 and 24 months, including minor gastrointestinal complaints and headache) can be expected. Second, diagnosis and treatment by primary-care physicians without additional resources are sustainable. This good news is clearly important for public-health planners in China. Coming hard on the heels of the finding of a 63% treatment gap,4 and thus the probable existence of more than five million untreated people
See Articles page 46
CNES, 2002 distribution spot image/Science Photo Library
For at least 25 years, the use of phenobarbital has declined substantially in developed nations because of persistent reports of adverse cognitive side-effects, especially hyperkinetic disorder and attention deficit disorder in children. Nevertheless, driven by economic realities that include costs per tablet up to ten times less than the cost of the next most affordable anticonvulsant, physicians in developing countries have continued to prescribe phenobarbital.1 The drug is also recommended as the first choice treatment for epilepsy in many national treatment guidelines from these countries—eg, in the essential drugs list of Zimbabwe, 2000—and in the formulary of the World Health Organization. Reliable evidence for the efficacy of phenobarbital in the developing world setting has been notably lacking, and it is here that Wang and co-workers,2 in reporting this issue on 1324 patients who completed (of 2455 who began) 2 years of treatment with the drug, achieve impressive results. The partnership of the Beijing Neurosurgical Institute and the Global Campaign against Epilepsy3 ensured rigorous assessment of the data with careful inclusion criteria. Wang and colleagues made two separate points in their report. First, good efficacy (26% of patients were seizure free and 72% had more than a 50% reduction in seizure frequency after 24 months) and few adverse effects (27% of patients
Epilepsy imposes a large economic burden on health-care systems in rural China
17
Dysphagia in acute stroke: a long-awaited trial See Articles page 31
16
Carnaby and colleagues1 describe in this issue of The Lancet Neurology the results of their single centre, singleblind, randomised controlled trial of a standard behavioural intervention for dysphagia after stroke as “encouraging”. I might even add the adjective exciting, because this trial has shown, for the first time, that an intervention can improve outcomes of patients with dysphagia. Dysphagia is more common in severe stroke, but is independently associated with poorer long-term outcomes in mild-to-moderate stroke.2 This trial indicates that a behavioural intervention not only improves recovery of swallowing, but also reduces the complications of dysphagia and improves long-term outcomes. This finding lends support to the notion that the association between dysphagia and poor outcomes is, at least in part, causal. Because almost half of all patients admitted to hospital with an acute stroke worldwide have dysphagia an effective treatment would be applicable to millions of patients each year, and therefore might have huge public-health benefits. However, before we get overly excited we must consider the limitations of this study, some of which the authors describe in their discussion. There are several examples, even in the stroke specialty, of small, single-centre, randomised trials reporting positive results that are then not confirmed by larger multicentre trials. Two examples are the effectiveness of percutaneous endoscopic gastrostomy feeding after acute stroke3,4 and leisure therapy.5,6 Many small trials that yield overly optimistic results have methodological failures that can introduce bias. The most common problems are: failure to enrol enough patients to ensure that randomisation produces treatment groups with similar baseline characteristics; failure to use secure methods of randomisation that prevent any foreknowledge of treatment allocation; failure to properly blind outcome assessments; failure to follow up all patients and to include all patients randomised in the primary analyses; failure to stipulate in advance what the primary outcome is so that the analysis becomes a data dredging operation, which is likely to produce false positive results; and failure to prespecify if any, or how many, interim analyses will be undertaken so that the trial is stopped simply when the chief investigator identifies a significant effect that they
think will make their trial acceptable to the editors of journals. Carnaby and colleagues have avoided most, if not all, of these failures. They used a secure randomisation system to enrol a respectable number of patients who were well described at baseline. They prespecified their sample size, which they achieved, and prespecified their primary and secondary outcomes, which were assessed largely blind to treatment allocation. They were rigorous in follow-up and accounted for all patients in their analysis. Moreover, they should be congratulated for describing their trial, their interventions, and their results in a way that would allow others to replicate their work. We now need more precise estimates of the effectiveness of the interventions and whether more intensive, and thus presumably more expensive, behavioural interventions are more effective than less intensive ones. We also need to determine whether this behavioural intervention, delivered by less specialist, skilful, or enthusiastic speech and language therapists working in many different environments, can achieve similar improvements in clinically relevant outcomes. Determining whether videofluoroscopy is an essential component or whether it could be dispensed with or replaced by other techniques, such as fibreoptic endoscopic assessment of swallowing, would also be useful. To reliably answer these questions will need many hundreds, or more likely thousands, of patients to be recruited into randomised trials. These patients might be enrolled in further single-centre trials, with similar methods to those described by Carnaby and colleagues, which could then be summarised in an informative meta-analysis. However, a family of large multicentre trials including a factorial design to address the questions of videofluoroscopy or alternatives would be a better way forward. I am therefore frustrated and somewhat disappointed that this trial has taken almost 5 years to report its results. These are 5 wasted years when further studies could have been planned, funding identified, and recruitment started. If we are to deliver better treatment to our patients we have to streamline our research efforts, avoiding such delays in reporting encouraging results and establishing confirmatory http://neurology.thelancet.com Vol 5 January 2006
Reflection and Reaction
studies. Hopefully, initiatives such as the Stroke Research Networks, funded through the UK Clinical Research Collaboration and Chief Scientist Office in Scotland, which aim to provide infrastructure to facilitate multicentre stroke trials, will help us to assess such promising interventions more quickly. Delays in introducing thrombolytic therapy as a standard treatment for myocardial infarction worldwide, which arose from failures to summarise existing data indicating its effectiveness and to carry out really definitive trials earlier, have been estimated to have killed thousands of patients.7 I hope that our delay in confirming the effectiveness of a standard behavioural intervention for stroke patients with dysphagia does not prove to be as costly. Martin Dennis
Division of Clinical Neurosciences, University of Edinburgh, Western General Hospital, Edinburgh EH4 2XU, UK [email protected]
See UK Clinical Research Collaboration home page: http://www.ukcrc.org/
I have no conflicts of interest. 1 2 3
4
5 6
7
Carnaby G, Hankey GJ, Pizzi J. Behavioural intervention for dysphagia in acute stroke: a randomised controlled trial. Lancet Neurol 2005; 5: 31–37. Smithard DG, O’Neill PA, Park C, et al. Complications and outcome after acute stroke: does dysphagia matter? Stroke 1996; 27: 1200–04. Norton B, Homer-Ward M, Donnelly MT, Long RG, Holmes GKT. A randomised prospective comparison of percutaneous endoscopic gastrostomy and nasogastric tube feeding after acute dysphagic stroke. BMJ 1996; 312: 13–16. FOOD Trial Collaboration. Effect of timing and method of enteral tube feeding for dysphagic stroke patients (FOOD): a multicentre randomised controlled trial. Lancet 2005; 365: 764–72. Drummond A, Walker MF. A randomised controlled trial of leisure therapy after stroke. Clin Rehab 1994; 9: 283–90. Parker CJ, Gladman JR, Drummond AE, et al. A multicentre randomized controlled trial of leisure therapy and conventional occupational therapy after stroke: Trial of Occupational Therapy and Leisure. Clin Rehab 2001; 15: 42–52. Lau J, Antman EM, Jimenez-Silva J, Kupelnick B, Mosteller F, Chalmers TC. Cumulative meta-analysis of therapeutic trials for myocardial infarction. New Engl J Med1992; 327: 248–54.
Phenobarbital for convulsive epilepsy at primary care level
http://neurology.thelancet.com Vol 5 January 2006
had one of all adverse effectss between 19 and 24 months, including minor gastrointestinal complaints and headache) can be expected. Second, diagnosis and treatment by primary-care physicians without additional resources are sustainable. This good news is clearly important for public-health planners in China. Coming hard on the heels of the finding of a 63% treatment gap,4 and thus the probable existence of more than five million untreated people
See Articles page 46
CNES, 2002 distribution spot image/Science Photo Library
For at least 25 years, the use of phenobarbital has declined substantially in developed nations because of persistent reports of adverse cognitive side-effects, especially hyperkinetic disorder and attention deficit disorder in children. Nevertheless, driven by economic realities that include costs per tablet up to ten times less than the cost of the next most affordable anticonvulsant, physicians in developing countries have continued to prescribe phenobarbital.1 The drug is also recommended as the first choice treatment for epilepsy in many national treatment guidelines from these countries—eg, in the essential drugs list of Zimbabwe, 2000—and in the formulary of the World Health Organization. Reliable evidence for the efficacy of phenobarbital in the developing world setting has been notably lacking, and it is here that Wang and co-workers,2 in reporting this issue on 1324 patients who completed (of 2455 who began) 2 years of treatment with the drug, achieve impressive results. The partnership of the Beijing Neurosurgical Institute and the Global Campaign against Epilepsy3 ensured rigorous assessment of the data with careful inclusion criteria. Wang and colleagues made two separate points in their report. First, good efficacy (26% of patients were seizure free and 72% had more than a 50% reduction in seizure frequency after 24 months) and few adverse effects (27% of patients
Epilepsy imposes a large economic burden on health-care systems in rural China
17