- Email: [email protected]
The Canadian CT Head Rule
CORRESPONDENCE
consequences of being overweight or obese and their roles in the progression of chronic nephropathies have probably been underestimated. Beyond the nephrotic proteinuria seen sometimes in massive obesity, functional and structural renal changes have been directly associated with excessive bodyweight in the early stages of obesity.3 Furthermore, obese patients are at risk of developing proteinuria and chronic renal failure after unilateral nephrectomy.4 We have shown, in a large cohort of patients with primary IgA nephropathy, that excessive bodyweight (conventionally defined as a bodymass index [BMI] >25 kg/m2) is independently associated with the amount of proteinuria at diagnosis and the severity of renal lesions on the first renal biopsy.5 Our findings suggested an independent contribution of BMI on the formation of vascular, tubular, and interstitial lesions, after adjustment for the concomitant presence of hyperlipidaemia and blood glucose concentration. Cox’s regression analysis showed that BMI is an additional and independent risk factor for arterial hypertension and the development of chronic renal failure.5 BMI represents a simple clinical parameter that may be readily measured and integrated into clinical practice to assess the risk for renal progression at the time of diagnosis. However, whether weight reduction might improve or delay renal disease progression remains to be prospectively assessed. *Fabrice Bonnet, François Berthezène, François Berthoux *Department of Nephrology, Dialysis and Renal Transplantation, University Hospital of Sainte-Etienne, Sainte-Etienne 42055, France; and Department of Endocrinology, University Hospital of Lyon 1
2
3
4
5
Ruggenenti P, Schieppati A, Remuzzi G. Progression, remission, regression of chronic renal diseases. Lancet 2001; 357: 1601–08. Mokdad AH, Serdula MK, Dietz WH, Bowman BA, Marks JS, Koplan JP. The spread of the obesity epidemic in the United States, 1991–1998. JAMA 1999; 282: 1519–22. Henegar JR, Bigler SA, Henegar LK, Tyagi SC, Hall JE. Functional and structural changes in the kidney in the early stages of obesity. J Am Soc Nephrol 2001; 12: 1211–17. Praga M, Hernandez E, Herrero JC, et al. Influence of obesity on the appearance of proteinuria and renal insufficiency after unilateral nephrectomy. Kidney Int 2000; 58: 2111–18. Bonnet F, Deprele C, Sassolas A, et al. Excessive body weight as a new independent risk factor for clinical and pathological progression in primary IgA nephritis. Am J Kidney Dis 2001; 37: 720–27.
THE LANCET • Vol 358 • September 22, 2001
Guardianship in research Sir—In your June 2 refractory item1 you address the issue of guardianship within the context of clinical trials and research. A local research ethics committee (LREC) might decree that any one patient should not be entered into more than three concurrent trials despite eligibility. Participation in clinical trials should not be imposed upon patients, but there is no documented evidence that involvement in multiple trials constitutes an excessive burden or is harmful to the interests of patients. Rather than maintaining that there is a moral obligation for patients to participate in one or more clinical trials, clinicians and researchers attempt to foster a spirit of shared ownership and responsibility within an ethical framework.2 Nonetheless, they should “protect themselves from rightism and ill-informed attacks from self-appointed armchair ethicists”.3 The rights and needs of individual patients are clearly paramount and must be balanced with the trial process and development of optimum treatment strategies. Through a process of education of patients and members of cancer teams, recruitment to clinical trials should increase. Nonetheless, ethicists must exercise a constructive and objective gate-keeping function; members of an LREC frequently deliberate on research proposals outside their immediate field of expertise. There is a risk that a balanced and fair judgment will not be reached if the demands of trial participation (frequent visits to hospital, additional blood tests and other investigations, disruption of work and family life, &c) dominate discussions and are perceived as overly intrusive. Such advocacy and protectionism might ultimately undermine and threaten an otherwise well designed trial that ensures adequate data collection and keeps to a minimum the chance of inconclusive results with lack of statistical significance. Perhaps patients themselves should be more involved in assessing the acceptability of the trial burden. To find out the psychosocial impact of involvement in multiple clinical trials, our group is designing a prospective qualitative research study to elicit the views and experiences of patients offered multiple trial entry, and to monitor their progress and wellbeing throughout and after participation in three or more clinical trials. J R Benson Cambridge Breast Unit, Addenbrooke’s Hospital, Cambridge CB2 2QQ, UK
1 2
3
Anon. Guardianship. Lancet 2001; 357: 1808. Thornton H. The voice of the breast cancer patient: a lonely cry in the wilderness. Eur J Cancer 1997; 33: 825–28. Baum M. The ethics of randomised controlled trials. Eur J Surg Oncol 1995; 21: 136–39.
The Canadian CT Head Rule Sir—Ian Stiell and colleagues (May 5, p 1391),1 in their study of use of computed tomography in patients with minor head injury, made every effort to follow up all the patients with head trauma to a clinical conclusion. Patients who did not have CT scans were followed up by telephone for 2 weeks to explore whether they had clinically important brain lesions. Of concern to us are the 363 patients who were enrolled initially, who did not get CT scans and could not be reached for the proxy outcome measures. If we assume the worst, in which these patients could not be reached because of death, coma, or amnesia (did they remember the way home?), then the results are changed strikingly. If we assume that all the lost patients had a no CT decision according to the Canadian CT Head Rule, sensitivity would then be 40·5%, and specificity would remain at 49·6%. A more reasonable concern might be that those 363 had clinically important brain injury at the same rate as those who completed the study (254 [8%] of 3121). If the decision rule predicted yes, as in the cohort who were followed up completely, the rule’s sensitivity becomes 93% and the specificity remains at around 50%. We are uncomfortable with assuming the worst and the average outcomes, since we have no clinical information about those patients. Did they resemble the rest of the cohort in cause of injury, use of alcohol, time to arrival at the accident and emergency department, amnesia, and so on? We look forward to resolving this minor question and congratulate Stiell and colleagues on a fascinating study. *Indi Maharaj, Lorraine Tosiello Overlook Hospital, Atlantic Health System, 99 Beauvoir Avenue, PO Box 220, Summit, NJ 07902, USA 1
Stiell IG, Wells GA, Vandemheen K, et al, for the CCC Study Group. The Canadian CT Head Rule for patients with minor head injury. Lancet 2001; 357: 1391–96.
Sir—Ian Stiell and colleagues’ study,1 although timely and appropriate in the current medical environment, is limited by its sample and design. The outcome measures were need
1013
For personal use. Only reproduce with permission from The Lancet Publishing Group.
CORRESPONDENCE
for neurological intervention and clinically important brain injury on CT, but, surprisingly, a third of patients did not actually undergo CT. Telephone follow-up, which included a neuropsychiatric test that has not been previously validated in studies on minor head injury was deemed equivalent to a negative CT. Because this proxy outcome measure did not identify 13% of clinically important brain injury, their use of it as a screening tool is concerning. In Canada, where only 30% of hospitals have a CT scanner, physicians are seeking ways to identify patients with minor head injury who should undergo scanning. In the USA, however, where we are criticised for overuse of our many CT scanners, physicians are seeking ways to identify patients who can safely forgo scanning. The first three items of Stiell and colleagues’ Canadian CT guidelines (Glasgow coma score <15, suspected open or depressed skull fracture, signs of basilar skull fracture) are well established indications for CT, and their results simply validate this. They intimate that widespread use of CT in minor head injury is limited to the USA, but reports from Scandinavia and Italy show a more global trend, emphasising the increasing need for guidelines to identify patients that can safely forgo CT.2,3 We have published such a set of guidelines. We used acute intracranial injury on CT as our outcome measure, because even small intracranial haematomas and isolated contusions typically warrant hospital admission in the USA.4 We included only patients with traumatic loss of consciousness or post-traumatic amnesia who had a Glasgow coma score of 15 and a normal brief neurological examination in the accident and emergency department. Stiell erroneously states that Haydel and colleagues’4 proposed guidelines are not reliable, sensitive, or specific enough to safely and efficiently guide clinicians in their use of CT. Application of our proposed guideline would have lessened the need for CT by 22% without missing any patients with an abnormal CT. We noted that, in the absence of headache, emesis, age older than 60 years, drug or alcohol intoxication, convulsion, trauma visible above the clavicles, and short-term memory deficits a CT is not needed. Our results were 100% sensitive and inarguably safe. In other countries, where CT is not readily available, implementation of our guidelines might increase the use of CT, but reports corroborate that
1014
this increase is necessary to identify all patients with intracranial injury.
to further assess sensitivity, specificity, reliability, and acceptability.
Micelle J Haydel Section of Emergency Medicine, Louisiana State University Health Science Center at New Orleans, 1532 Tulane Avenue-Charity Hospital 13th Floor, New Orleans, LA 70112, USA (e-mail: [email protected]) 1
2
3
4
Stiell IG, Wells GA, Vandemheen K, et al, for the CCC Study Group. The Canadian CT Head Rule for patients with minor head injury. Lancet 2001; 357: 1391–96. Ingebrigtsen T, Romner B, Kock-Jensen C. Scandinavian guidelines for initial management of minimal, mild, and moderate head injuries: the Scandinavian Neurotrauma Committee. J Trauma 2000; 48: 760–66. The Study Group on Head Injury of the Italian Society for Neurosurgery. Guidelines for minor head injured patients’ management in adult age. J Neurosurg Sci 1996; 40: 11–15. Haydel MJ, Preston CA, Mills TJ, Luber S, Blaudeau E, DeBlieux PM. Indications for computed tomography in patients with minor head injury. N Engl J Med 2000; 343: 100–05.
Authors’ reply Sir—We understand the concern of Indi Maharaj and Lorraine Tosiello that the patients not reached in followup might have harboured serious injuries, but believe that we can completely allay any such concerns. First, we did a derivation study and not a validation study; hence we only assessed patients for whom there were complete outcome data apart from their CT findings and follow-up neurological status. Second, the patients who did not undergo CT were those who were completely intact neurologically and who were felt to be at extremely low risk by the experienced clinicians treating them. Third, the study sites were the primary neurosurgical centres for the respective cities and we would have seen such patients return if there had been adverse outcomes. No patient returned. Hence, we are very comfortable that the sensitivity of the derived Canadian CT Head Rule is 100% for need of neurological intervention. We regret that our report may have cast Micelle Haydel’s study in a more negative light than we had intended and we apologise for that. We do believe that the sensitivity of the New Orleans criteria for predicting need for neurological intervention has not been well established because Haydel’s study had only six patients who required surgery. We also continue to be concerned that the New Orleans criteria have unnecessarily low specificity. Hence, we stand by our opinion that the New Orleans criteria should undergo explicit and prospective assessment in multiple sites
*Ian G Stiell, George A Wells, for the CCC Study Group *Clinical Epidemiology unit, F6, and Departments of Epidemiology and Community Medicine, Ottawa Hospital Loeb Health Research Institute, Ottawa, Ontario K1Y 4E9, Canada
Adequacy of treatment in terminally ill patients Sir—Stefan Weiss and colleagues (April 28, p 1311)1 report that the number of inadequately treated patients among terminally ill patients is not as large as perceived, based on the fact that most patients are willing to tolerate pain. In reaction, Sam Hjelmeland Ahmedzai in his April 28 commentary2 argues that patients choose pain because the alternative of troublesome side-effects is even worse. Weiss’ study is unique because of the large number of patients from various diagnostic groups. I suggest, however, that Weiss’ conclusion might be based on a misinterpretation of results. First, the use of patients’ satisfaction alone to monitor the quality of pain treatment is dubious. Pain treatment in terminally ill patients is complex because of the divergent components that play a part, such as the nature of the pain, the pain treatment itself, and the patients’ reaction to analgesics. Besides the use of patients’ judgment about pain, professionally-oriented outcome measures should be used to assess pain treatment. No empirical data were presented on the use of analgesics (name, dose, frequency, and route) and severity of side-effects. To focus on only one patient-oriented outcome measure is insufficient. Second, patients’ satisfaction is not a valid and reliable measure.3 The socalled high pain high satisfaction paradox occurs with patients being satisfied with the pain treatment despite high-degree intensity. In a study in patients with chronic cancer pain, we noted that 60% were treated inadequately for their pain,4 whereas most hesitated to use medication and waited too long before asking for pain relief.5 Most patients thought that cancer-pain patients are overmedicated and that addiction is inevitable. The mean pain-knowledge score was 55 on a scale of 0–100. This dissonance arises when patients do not know that pain relief is achievable in most patients. Third, a multidisciplinary team of clinicians, as mentioned by Ahmedzai, will not solve the issue of patients’ fears. At a time when patients’ length of stay in the hospital is decreasing and emphasis
THE LANCET • Vol 358 • September 22, 2001
For personal use. Only reproduce with permission from The Lancet Publishing Group.
consequences of being overweight or obese and their roles in the progression of chronic nephropathies have probably been underestimated. Beyond the nephrotic proteinuria seen sometimes in massive obesity, functional and structural renal changes have been directly associated with excessive bodyweight in the early stages of obesity.3 Furthermore, obese patients are at risk of developing proteinuria and chronic renal failure after unilateral nephrectomy.4 We have shown, in a large cohort of patients with primary IgA nephropathy, that excessive bodyweight (conventionally defined as a bodymass index [BMI] >25 kg/m2) is independently associated with the amount of proteinuria at diagnosis and the severity of renal lesions on the first renal biopsy.5 Our findings suggested an independent contribution of BMI on the formation of vascular, tubular, and interstitial lesions, after adjustment for the concomitant presence of hyperlipidaemia and blood glucose concentration. Cox’s regression analysis showed that BMI is an additional and independent risk factor for arterial hypertension and the development of chronic renal failure.5 BMI represents a simple clinical parameter that may be readily measured and integrated into clinical practice to assess the risk for renal progression at the time of diagnosis. However, whether weight reduction might improve or delay renal disease progression remains to be prospectively assessed. *Fabrice Bonnet, François Berthezène, François Berthoux *Department of Nephrology, Dialysis and Renal Transplantation, University Hospital of Sainte-Etienne, Sainte-Etienne 42055, France; and Department of Endocrinology, University Hospital of Lyon 1
2
3
4
5
Ruggenenti P, Schieppati A, Remuzzi G. Progression, remission, regression of chronic renal diseases. Lancet 2001; 357: 1601–08. Mokdad AH, Serdula MK, Dietz WH, Bowman BA, Marks JS, Koplan JP. The spread of the obesity epidemic in the United States, 1991–1998. JAMA 1999; 282: 1519–22. Henegar JR, Bigler SA, Henegar LK, Tyagi SC, Hall JE. Functional and structural changes in the kidney in the early stages of obesity. J Am Soc Nephrol 2001; 12: 1211–17. Praga M, Hernandez E, Herrero JC, et al. Influence of obesity on the appearance of proteinuria and renal insufficiency after unilateral nephrectomy. Kidney Int 2000; 58: 2111–18. Bonnet F, Deprele C, Sassolas A, et al. Excessive body weight as a new independent risk factor for clinical and pathological progression in primary IgA nephritis. Am J Kidney Dis 2001; 37: 720–27.
THE LANCET • Vol 358 • September 22, 2001
Guardianship in research Sir—In your June 2 refractory item1 you address the issue of guardianship within the context of clinical trials and research. A local research ethics committee (LREC) might decree that any one patient should not be entered into more than three concurrent trials despite eligibility. Participation in clinical trials should not be imposed upon patients, but there is no documented evidence that involvement in multiple trials constitutes an excessive burden or is harmful to the interests of patients. Rather than maintaining that there is a moral obligation for patients to participate in one or more clinical trials, clinicians and researchers attempt to foster a spirit of shared ownership and responsibility within an ethical framework.2 Nonetheless, they should “protect themselves from rightism and ill-informed attacks from self-appointed armchair ethicists”.3 The rights and needs of individual patients are clearly paramount and must be balanced with the trial process and development of optimum treatment strategies. Through a process of education of patients and members of cancer teams, recruitment to clinical trials should increase. Nonetheless, ethicists must exercise a constructive and objective gate-keeping function; members of an LREC frequently deliberate on research proposals outside their immediate field of expertise. There is a risk that a balanced and fair judgment will not be reached if the demands of trial participation (frequent visits to hospital, additional blood tests and other investigations, disruption of work and family life, &c) dominate discussions and are perceived as overly intrusive. Such advocacy and protectionism might ultimately undermine and threaten an otherwise well designed trial that ensures adequate data collection and keeps to a minimum the chance of inconclusive results with lack of statistical significance. Perhaps patients themselves should be more involved in assessing the acceptability of the trial burden. To find out the psychosocial impact of involvement in multiple clinical trials, our group is designing a prospective qualitative research study to elicit the views and experiences of patients offered multiple trial entry, and to monitor their progress and wellbeing throughout and after participation in three or more clinical trials. J R Benson Cambridge Breast Unit, Addenbrooke’s Hospital, Cambridge CB2 2QQ, UK
1 2
3
Anon. Guardianship. Lancet 2001; 357: 1808. Thornton H. The voice of the breast cancer patient: a lonely cry in the wilderness. Eur J Cancer 1997; 33: 825–28. Baum M. The ethics of randomised controlled trials. Eur J Surg Oncol 1995; 21: 136–39.
The Canadian CT Head Rule Sir—Ian Stiell and colleagues (May 5, p 1391),1 in their study of use of computed tomography in patients with minor head injury, made every effort to follow up all the patients with head trauma to a clinical conclusion. Patients who did not have CT scans were followed up by telephone for 2 weeks to explore whether they had clinically important brain lesions. Of concern to us are the 363 patients who were enrolled initially, who did not get CT scans and could not be reached for the proxy outcome measures. If we assume the worst, in which these patients could not be reached because of death, coma, or amnesia (did they remember the way home?), then the results are changed strikingly. If we assume that all the lost patients had a no CT decision according to the Canadian CT Head Rule, sensitivity would then be 40·5%, and specificity would remain at 49·6%. A more reasonable concern might be that those 363 had clinically important brain injury at the same rate as those who completed the study (254 [8%] of 3121). If the decision rule predicted yes, as in the cohort who were followed up completely, the rule’s sensitivity becomes 93% and the specificity remains at around 50%. We are uncomfortable with assuming the worst and the average outcomes, since we have no clinical information about those patients. Did they resemble the rest of the cohort in cause of injury, use of alcohol, time to arrival at the accident and emergency department, amnesia, and so on? We look forward to resolving this minor question and congratulate Stiell and colleagues on a fascinating study. *Indi Maharaj, Lorraine Tosiello Overlook Hospital, Atlantic Health System, 99 Beauvoir Avenue, PO Box 220, Summit, NJ 07902, USA 1
Stiell IG, Wells GA, Vandemheen K, et al, for the CCC Study Group. The Canadian CT Head Rule for patients with minor head injury. Lancet 2001; 357: 1391–96.
Sir—Ian Stiell and colleagues’ study,1 although timely and appropriate in the current medical environment, is limited by its sample and design. The outcome measures were need
1013
For personal use. Only reproduce with permission from The Lancet Publishing Group.
CORRESPONDENCE
for neurological intervention and clinically important brain injury on CT, but, surprisingly, a third of patients did not actually undergo CT. Telephone follow-up, which included a neuropsychiatric test that has not been previously validated in studies on minor head injury was deemed equivalent to a negative CT. Because this proxy outcome measure did not identify 13% of clinically important brain injury, their use of it as a screening tool is concerning. In Canada, where only 30% of hospitals have a CT scanner, physicians are seeking ways to identify patients with minor head injury who should undergo scanning. In the USA, however, where we are criticised for overuse of our many CT scanners, physicians are seeking ways to identify patients who can safely forgo scanning. The first three items of Stiell and colleagues’ Canadian CT guidelines (Glasgow coma score <15, suspected open or depressed skull fracture, signs of basilar skull fracture) are well established indications for CT, and their results simply validate this. They intimate that widespread use of CT in minor head injury is limited to the USA, but reports from Scandinavia and Italy show a more global trend, emphasising the increasing need for guidelines to identify patients that can safely forgo CT.2,3 We have published such a set of guidelines. We used acute intracranial injury on CT as our outcome measure, because even small intracranial haematomas and isolated contusions typically warrant hospital admission in the USA.4 We included only patients with traumatic loss of consciousness or post-traumatic amnesia who had a Glasgow coma score of 15 and a normal brief neurological examination in the accident and emergency department. Stiell erroneously states that Haydel and colleagues’4 proposed guidelines are not reliable, sensitive, or specific enough to safely and efficiently guide clinicians in their use of CT. Application of our proposed guideline would have lessened the need for CT by 22% without missing any patients with an abnormal CT. We noted that, in the absence of headache, emesis, age older than 60 years, drug or alcohol intoxication, convulsion, trauma visible above the clavicles, and short-term memory deficits a CT is not needed. Our results were 100% sensitive and inarguably safe. In other countries, where CT is not readily available, implementation of our guidelines might increase the use of CT, but reports corroborate that
1014
this increase is necessary to identify all patients with intracranial injury.
to further assess sensitivity, specificity, reliability, and acceptability.
Micelle J Haydel Section of Emergency Medicine, Louisiana State University Health Science Center at New Orleans, 1532 Tulane Avenue-Charity Hospital 13th Floor, New Orleans, LA 70112, USA (e-mail: [email protected]) 1
2
3
4
Stiell IG, Wells GA, Vandemheen K, et al, for the CCC Study Group. The Canadian CT Head Rule for patients with minor head injury. Lancet 2001; 357: 1391–96. Ingebrigtsen T, Romner B, Kock-Jensen C. Scandinavian guidelines for initial management of minimal, mild, and moderate head injuries: the Scandinavian Neurotrauma Committee. J Trauma 2000; 48: 760–66. The Study Group on Head Injury of the Italian Society for Neurosurgery. Guidelines for minor head injured patients’ management in adult age. J Neurosurg Sci 1996; 40: 11–15. Haydel MJ, Preston CA, Mills TJ, Luber S, Blaudeau E, DeBlieux PM. Indications for computed tomography in patients with minor head injury. N Engl J Med 2000; 343: 100–05.
Authors’ reply Sir—We understand the concern of Indi Maharaj and Lorraine Tosiello that the patients not reached in followup might have harboured serious injuries, but believe that we can completely allay any such concerns. First, we did a derivation study and not a validation study; hence we only assessed patients for whom there were complete outcome data apart from their CT findings and follow-up neurological status. Second, the patients who did not undergo CT were those who were completely intact neurologically and who were felt to be at extremely low risk by the experienced clinicians treating them. Third, the study sites were the primary neurosurgical centres for the respective cities and we would have seen such patients return if there had been adverse outcomes. No patient returned. Hence, we are very comfortable that the sensitivity of the derived Canadian CT Head Rule is 100% for need of neurological intervention. We regret that our report may have cast Micelle Haydel’s study in a more negative light than we had intended and we apologise for that. We do believe that the sensitivity of the New Orleans criteria for predicting need for neurological intervention has not been well established because Haydel’s study had only six patients who required surgery. We also continue to be concerned that the New Orleans criteria have unnecessarily low specificity. Hence, we stand by our opinion that the New Orleans criteria should undergo explicit and prospective assessment in multiple sites
*Ian G Stiell, George A Wells, for the CCC Study Group *Clinical Epidemiology unit, F6, and Departments of Epidemiology and Community Medicine, Ottawa Hospital Loeb Health Research Institute, Ottawa, Ontario K1Y 4E9, Canada
Adequacy of treatment in terminally ill patients Sir—Stefan Weiss and colleagues (April 28, p 1311)1 report that the number of inadequately treated patients among terminally ill patients is not as large as perceived, based on the fact that most patients are willing to tolerate pain. In reaction, Sam Hjelmeland Ahmedzai in his April 28 commentary2 argues that patients choose pain because the alternative of troublesome side-effects is even worse. Weiss’ study is unique because of the large number of patients from various diagnostic groups. I suggest, however, that Weiss’ conclusion might be based on a misinterpretation of results. First, the use of patients’ satisfaction alone to monitor the quality of pain treatment is dubious. Pain treatment in terminally ill patients is complex because of the divergent components that play a part, such as the nature of the pain, the pain treatment itself, and the patients’ reaction to analgesics. Besides the use of patients’ judgment about pain, professionally-oriented outcome measures should be used to assess pain treatment. No empirical data were presented on the use of analgesics (name, dose, frequency, and route) and severity of side-effects. To focus on only one patient-oriented outcome measure is insufficient. Second, patients’ satisfaction is not a valid and reliable measure.3 The socalled high pain high satisfaction paradox occurs with patients being satisfied with the pain treatment despite high-degree intensity. In a study in patients with chronic cancer pain, we noted that 60% were treated inadequately for their pain,4 whereas most hesitated to use medication and waited too long before asking for pain relief.5 Most patients thought that cancer-pain patients are overmedicated and that addiction is inevitable. The mean pain-knowledge score was 55 on a scale of 0–100. This dissonance arises when patients do not know that pain relief is achievable in most patients. Third, a multidisciplinary team of clinicians, as mentioned by Ahmedzai, will not solve the issue of patients’ fears. At a time when patients’ length of stay in the hospital is decreasing and emphasis
THE LANCET • Vol 358 • September 22, 2001
For personal use. Only reproduce with permission from The Lancet Publishing Group.