| | Effect of an Integrated Care Pathway on Acute Asthma/Wheeze in Children Attending Hospital: Cluster Randomized TrialReceived 7 April 2007; received in revised form 13 August 2007; accepted 17 September 2007. published online 19 November 2007.
|
Refers to article:
|
|
Reliable Implementation of Clinical Pathways: What Will It Take—That Is the Question
Jeffrey M. Simmons, Uma R. Kotagal
The Journal of Pediatrics
March 2008 (Vol. 152, Issue 3, Pages 303-304)
Full Text |
Full-Text PDF (63 KB)
|
ObjectiveTo determine whether an integrated care pathway (ICP) could improve care delivered to patients coming to an emergency department only or to patients who were subsequently admitted. Study designChildren (age, 2-16 years; n = 298) coming to the ED with acute asthma/wheeze, were randomized by using a cluster design to either standard care or care delivered by an ICP. ResultsChildren discharged from the ED who received care with an ICP (n = 118) received more prednisolone (81%; standard, 63%; P = .03) and increased advice to obtain primary care review (72%; standard, 33%; P < .0001). A total of 180 children were admitted (94 ICP, 86 standard). The rate of recovery was unchanged by ICP. The mean ICP length of stay (37.6 hours; range, 33.5-42.4 hours), was 93% of the mean standard length of care (40.7 hours; range, 35.9-46; P = .36). When a discharge checklist was completed (60 of 94 cases), the mean ICP length of stay was 34.2 hours (range, 30.5-38.4 hours; P = .07 versus standard). An ICP resulted in a 30% reduction in prescribing errors (mean, 10.4; standard, 14.8; P = .002). Eighty-four of 94 children with an ICP received a 48-hour discharge plan (89%) versus 35 of 86 children with standard care (41%). More clinical contacts were observed in children receiving care by an ICP (mean, 22, versus standard, 19.2: P = .0004). ConclusionAn acute asthma/wheeze ICP improved education and prescribing errors, modestly reduced the length of stay when discharge criteria were adhered to, but did not influence recovery time. Further consideration of the effect on staff workload is required. Asthma affects 5.2 million people in the United Kingdom, including 1.8 million children, and it is a significant burden on patients, families, and healthcare systems. Initiatives to improve asthma care focus on evidence-based guidelines and reducing unscheduled admissions to hospital through education. Poor implementation of guidelines1, 2 and a dependence on additional (usually human) resources to deliver successful educational programs3 have limited the success of such initiatives. Poor formatting of guideline recommendations to enable use during everyday clinical processes can significantly impede their adoption into daily practice.4, 5 Appropriate design may help streamline management plans across handovers,6 reduce unnecessary variability in care,4 and prompt clinicians to the best evidence-based practice,7 whatever their level of experience. Integrated care pathways (ICPs) are structured multidisciplinary documents that direct care (incorporating guideline recommendations), reducing unnecessary clinical variation.8, 9 Although there have been a number of studies assessing the introduction of acute asthma/wheeze ICPs in hospital-based care, they generally have been retrospective comparisons10, 11, 12, 13, 14 or studies of insufficient power.15 Our trial incorporates a cluster design in 7-day blocks to facilitate the observation of professional behavior with the 2 types of clinical documentation. Randomizing the whole acute admission unit to 1 category of documentation during each cluster was preferred to minimize the potential risk of Hawthorne effect (as randomizing individuals might) or introducing confounders (as randomization of individual patients to separate wards [ICP or standard] with different skill mixes might). Patients were exposed to only 1 form of documentation during each cluster period (continued for the duration of their stay) in 7-day periods to accommodate the variability of pediatric asthma hospitalizations. Our aim was to assess the effect of the introduction of an ICP for acute asthma on the length of stay and rate of recovery in children admitted to the hospital with an acute asthma exacerbation. Important secondary outcomes were whether the pathway could improve the education provided to parents, increase knowledge of need for subsequent review in primary care, and reduce the potential for clinical error (prescribing errors) without significantly increasing workload to staff. Methods  Subjects The study was conducted in the emergency department (ED) and the medical wards of a pediatric university hospital in Edinburgh, Scotland, UK. All patients (aged 2-16 years inclusively) who came to the hospital ED with acute asthma/wheeze were eligible for inclusion. Children with acute viral bronchiolitis (fine crackles and wheeze on auscultation) were excluded from the study, as were children requiring intensive care (intravenous bronchodilator) or children with significant cardiovascular or neurological deficit. Verbal consent was obtained for children discharged home from the ED, and written consent was obtained for children who were admitted. Lothian Local Research Ethics Committee 03, UK, gave ethical approval for the study. Study Design The study was a cluster-randomized trial during a 26-week period (Aug 23, 2004-Feb 15, 2005) of standard care or care delivered with an ICP. Patients received care according to permuted block cluster randomization (7-day periods in blocks of 8 weeks) generated before the start of the study (R.P.). All children who came to the ED during the study period were assessed. Allocation of clusters was not concealed. Seven-day clusters began on Monday morning at 9:00 a.m., and finished on Monday morning at 8:59 a.m. ED triage nurses dispensed the appropriate documentation to staff during each cluster period. Standard care provided separate documentation for nursing, medical, clinical observation, and prescribing charts. There were no prompts within the standard documentation for timing of decisions or discharge and no guidance on education that should be provided. An ICP combines all nursing, medical, clinical observation, and prescribing charts chronologically within a single document. A multidisciplinary team consisting of medical, nursing, and pharmacy members developed the ICP, incorporating evidence from Cochrane Reviews and the SIGN/BTS asthma guideline.16 Sections prompted the identification of discharge criteria and directed education and issue of action plans. The study assessed the introduction of the ICP in a pragmatic, real life way in which opportunities for ongoing education would be limited. Routine annual education is provided to all acute receiving staff about the asthma education materials noted in the study (use of peak flow, inhaler techniques, emergency medical plans, etc). Tutorials on the use of the ICP were provided to all staff groups only in the month before the introduction of the ICP or to new members of staff arriving during the study, but no further teaching was offered. Outcome Measures Children admitted to the hospital formed the core group for the study. Information from the ED was included, with data on the speed of recovery of physiological variables (heart rate, respiratory rate, oxygen requirement), quantity and speed of reduction for bronchodilator requirement, time to fulfilling discharge criteria, education provided, and prescribing errors. Data collection (C.L.) was cross-checked for consistency in the first 50 patients (A.M.). The study group set predefined definitions of prescribing errors. Prescribing errors were classified in sections according to good prescribing practice for dosing, administration, clarity, legal issues, and other. Further subdivision was made in minor, moderate, and major errors.17 The overall number of prescribing errors per patient was compared. In addition, we evaluated the relative change in minor, moderate, and major errors resulting from the introduction of the ICP to gauge whether expected improvements in minor prescribing errors would be reflected by improvements in moderate and major errors. Two study members (C.L., M.D.) independently assessed prescribing records with discrepancies resolved by agreement (C.L., M.D., S.C.). Parents of children admitted to the hospital were requested to complete a questionnaire and were also contacted by telephone at 14 days after discharge from the hospital with a standardized scripted telephone questionnaire. Telephone questions could be answered spontaneously or recalled after 1 subsequent non-leading prompt. Analysis of Data The primary outcome was difference in length of hospital stay for the 2 cluster groups (from arrival at the ED to time discharged from the ward). In a retrospective audit, the average length of stay was 41 hours (SD, 15 hours). It was envisaged that the potential efficiencies of the ICP could reduce the length of stay by 20% (8 hours), because US-based studies had reported 25% to 49% reductions. An 80% power at a 2-sided 0.05 significance level would require 114 patients. We had no indication of the likely size of intracluster correlation and allowed a conservative increase of 50% to the standard sample size to allow for clustering effects. With an expected average cluster size of 7, this allows for intracluster correlation as high as 0.08. We therefore aimed to recruit 180 patients admitted to the hospital (90 to each limb). Analysis of the trial allowed for the clustering effects inherent in the trial design and for the possibility of children entering the trial more than once by the use of mixed models.18 SAS software version 9 was used for analysis, with the SAS procedure MIXED used for data assumed to follow a normal distribution and the procedure GLIMMIX used for binary data. Fitting the clusters (weeks) and individual children as random allows for within-cluster and within-children correlation. Observations made in successive periods were analyzed as repeated measures with a Toeplitz covariance structure. Continuous variables were either analyzed without transformation or after a logarithmic transformation. In situations in which agreement to normality was poor, a check on the validity of the conclusions was made by performing a non-parametric analysis, ignoring the clustering. Results A total of 322 ED visits of children aged 2 to 16 years with acute asthma/wheeze were made during the study period, with 298 included in the study (Figure). A total of 136 children (mean age, 5.7 years; SD, 2.9 years) came to the ED for the first time during ICP weeks, and 115 children (mean age, 5.8 years; SD, 3.2 years) in standard care weeks. Discharged Home from the ED In 118 visits, the child was discharged home after initial treatment, 69 of 163 cases in which care was provided with an ICP (42%) and 49 of 135 cases in which care was standard (36%; P = .38). Significant improvements in discharge prescriptions and education were achieved without a significant increase in time in ED or number of clinical contacts (Table I). Admissions to Hospital During the study, 180 ED visits (94 ICP, 86 standard) required that the child be admitted to hospital (159 children, 84 with ICP care, 75 with standard care). Parent questionnaires (Table II), were completed by 74 parents of children in the ICP group and 67 parents of children in the standard group who were admitted to the hospital for the first time, and these demonstrate very good comparability of the study groups at baseline. Primary Outcome Measures Patient recovery There was no difference in the groups in the rate of improvement in heart rate or respiratory rate or in the dose of bronchodilator prescribed for the first 24 hours (tests of interaction between treatment and time and tests for main effects of treatment, all P > .2; Table III; available at www.jpeds.com). The geometric mean time from arrival at the ED to achieve an oxygen saturation ≥94% was 9.3 hours (95% CI, 8.2-10.6 hours) for the ICP group and 10.3 hours (95% CI, 9-11.8 hours) for the standard care group (P = .3). The time to achieve 4-hour spacing of multidose spacer-delivered salbutamol (1000 mg) was not significantly different in the 2 groups (mean, 23.1 hours ICP [95% CI, 20.5-25.7]; mean, 23.6 hours standard [95% CI, 20.8-26.3]; P = .83). Discharge from hospital ward Introduction of the ICP was associated with lengths of stay that were 93% of those in standard care weeks (95% CI, 78%-110%; P = .36). The geometric mean for ICP was 37.6 hours (95% CI, 33.5-42.4), and for standard care it was 40.7 hours (95% CI, 35.9-46). The intracluster correlation coefficient for this primary outcome variable was 0.004. The protocol also pre-specified an analysis including age, sex, heart rate, respiratory rate, and oxygen saturation on arrival as co-variates. This gave almost identical results for the estimate of the treatment effect, but with a slightly larger intracluster correlation coefficient of 0.02. The ICP provided a clinical criteria checklist, which, once completed, authorized the discharge of children (without waiting for a ward round/senior medical review). The section required a date and time when all criteria had been fulfilled. This information was correctly documented in 60 patients (64% of ICP admissions). Subanalysis of this group identified the geometric mean length of time (from arrival at the ED) in which all discharge criteria were documented at 31.8 hours (95% CI, 28.4-35.6) with a corresponding length of stay of 34.2 hours (95% CI, 30.5-38.4). This compares with a length of stay in those patients in the ICP group in which the discharge criteria were not completed of 41.9 hours (95% CI, 32.7-53.7). In relative terms, the length of stay in these “per protocol” subjects was 85% (95% CI, 71-101; P = .07) of the corresponding stay in the standard care arm. To assess whether the size of any treatment effect on the length of the hospital stay depended on how long the trial had been running, a model was fitted including time since randomization as a continuous variable, with the treatment effect and a treatment by time interaction. The interaction term was not significant (P = .98), nor was the time effect (P =.81), indicating stability of the size of the treatment effect throughout the trial and an absence of any trend in the duration of hospital stay. There were fewer additional visits by children whose first visit was in the ICP weeks (16/136, 12%) compared with children whose first attendance was in standard care weeks (19/115, 17%), although the difference was not statistically significant (odds ratio, 0.6; 95% CI, 0.2-1.5; P = .27). Secondary Outcome Measures Provision of education and information to parents An admission to the hospital presents an important opportunity to provide education and information to parents (and children) on acute asthma/wheeze and its management. Introduction of the ICP had a significant effect on the documented amount of information provided by staff to parents/children during their stay in hospital (Table IV). Prescribing errors The ICP was associated with a 30% reduction in the total number of prescribing errors (mean, 14.8 for standard care, 10.4 for ICP; P = .002; Table V; available at www.jpeds.com). This effect was most significant for dosing and drug administration errors (approximately 6 times higher in standard care weeks). Legal errors were more prevalent in the ICP group, relating predominantly to poor documentation of the date on the prescribing page. In most cases, prescribing errors were minor, with relatively few moderate (mean, 3.4 both groups) or major errors (mean, 0.19 ICP versus 0.13 standard), and no statistically significant difference in the 2 study groups in these categories (moderate P = .95; major P = .33). Clinician workload The mean number of documented clinical contacts with the patient during the first 12 hours (including time in the ED) was higher in the ICP group for both medical and nursing staff (medical contacts: ICP 6, standard 5.5, P = .04; nursing contacts: ICP 16, standard 13.8, P = .002; all contacts: ICP 22, standard 19.2, P = .0004). Parental phone call at 14 days (standard script) A total of 133 parents could be contacted 14 days after admission (ICP 69/94, 73%; standard 64/86, 74%). A significantly higher proportion of parents contacted in the ICP group was documented as receiving advice on acute post discharge care and primary care follow-up. Although recall of this information was the same percentage in both groups at 14 days, the higher number of parents provided with this information in the ICP group lead to a doubling of the absolute number recalling advice to attend primary care post admission in the this group (Table VI). Adverse events No adverse events associated with introduction of the ICP were experienced. Discussion Use of an integrated care pathway for the children with acute asthma/wheeze who were admitted to the hospital was associated with a modest reduction in length of stay, fewer prescribing errors, provision of more education, and improved advice to attend primary care, although more clinical contacts were required during the patient stay. The ICP did not reduce the time spent in the ED or increase the speed of recovery from the acute asthma/wheeze exacerbation. In those patients discharged from the ED, significant improvements were made in the number provided with an adequate course of oral corticosteroids. The study demonstrates that multifaceted improvements in clinical care can be achieved without additional resources through structured redesign complementing the current care process.19 The reduction in length of stay afforded by the ICP (our primary outcome) was modest, but when clinicians completed discharge checklists, a process for an appreciably earlier discharge was initiated, such that children for whom the checklist was completed were discharged 7.7 hours earlier. Children for whom the checklist was not completed were discharged at a time similar to children in standard care weeks. This may suggest that our estimate of effect for the ICP was not unduly hopeful, had the ICP been used as designed. Other factors also influenced clinicians. In the ICP group, 26% children met discharge criteria between midnight and 9:00 a.m., but none were discharged within that time. Comparison with earlier asthma ICP studies is difficult because of variability in study quality and reporting. One randomized controlled study demonstrated a reduced length of stay in the ICP group, but it had insufficient power because of study design and had poorly matched groups at baseline.15 The study did not improve subsequent follow-up or medication provided. All other studies to date have been based in North America and have been non-randomized with retrospective control subjects.10, 11, 12, 13, 14 The comparative length of stay in all earlier pediatric studies was greater than the 41 hours of our standard care group (48, 54, 70, 71, 101 hours)10, 11, 12, 13, 15 and remained >41 hours in all studies except 1 after the introduction of the ICP. Comparisons of healthcare are complex, but this notable international difference in length of stay for a routine pediatric respiratory admission warrants further assessment. The principal mechanism for the success of the ICP in providing more education during the admission appeared to be prompts within the ICP, because the same resource materials were available to both groups. Providing education to children with acute asthma and their parents is effective in reducing future healthcare use,3, 20, 21 although studies to date (based in hospitals) have required additional and dedicated human resources.22, 23 Our study identifies that routine staff can provide a comprehensive package of education when prompted to do so at appropriate times during the patient stay. Earlier studies have suggested that asthma information in this format could be effectively provided by trained ward-based nurses,22 although a direct comparison with asthma educator advice has not yet been made. Printed prescribing sheets and relevant dose regimes present within the ICP reduced prescribing errors. The ICP did not have an impact on major prescribing errors, although the numbers in this category were small (and consistent with earlier reports).24 Whether a reduction of minor errors and education on good prescribing practice promoted by the ICP may foster a culture of safe prescribing that would reduce future risk of major (and clinically significant) prescribing errors remains to be seen. This was an open, unblinded study, and as such, it has inherent limitations. The Hawthorne effect may have been expected, but did not appear to occur for length of stay with our open label design. There was no facility within the study for staff to demonstrate proof of education being provided, merely a notation within the pathway that it had been provided. It is possible that staff noted education that was not provided, although the variable responses provided by staff to the prompts (Table I, Table IV), particularly to smoking advice and use of PEFR (peak expiratory flow rate), would suggest that these areas of the ICP were not treated simply as check lists. Also, the inability to blind observers scoring clinical notes may have introduced bias to the results; however, we attempted to limit this with the use of pre-specified data entry sheets and criteria for prescribing errors with cross checking. Our data on time to discharge (in which children with no completed ICP checklist were discharged effectively at the same time as patients in the standard group) and the significantly worse legal errors in ICP documentation may suggest that any potential observer bias was limited by the checks put in place. The failure by some staff to complete the discharge criteria checklist in ICP weeks appeared to have an important effect on the study outcome. There was no compulsion to use all aspects of the ICP, and use of this type of documentation is a cultural change in UK-based healthcare. Although some clinicians can be engaged to rapidly assimilate change,19 identifying those who are less adaptive and creating an environment to limit non-compliance will require refinement of the process to discourage non-compliance. Our study suggests that the rewards to patients, in earlier discharge and improved education opportunities, associated with better adherence of clinical staff to the ICP should warrant initiatives to encourage their use. The ICP used in this study could readily be adapted to other secondary care settings for adults and children with acute asthma/wheeze and, with further research, extend to the follow-up review of patients in primary care. ICPs are ideal vehicles to incorporate widely available guideline recommendations. With the current strains on medical education, increased handovers of medical and nursing information, and a focus on providing clearly documented consistent care, ICPs seem able to provide an adaptable, equitable, and effective level of care while remaining patient centered. Alan Fischer helped devise the asthma ICP; Neil Richardson helped devise pharmacy error rules. Information about the ICP may be obtained from the corresponding author. References 1. 1Grimshaw JM, Thomas RE, MacLennan G, Fraser C, Ramsay CR, Vale L, et al. Effectiveness and efficiency of guideline dissemination and implementation strategies. Health Technol Assess. 2004;8:iii-72. 2. 2Pearson MG, Ryland I, Harrison BD. Comparison of the process of care of acute severe asthma in adults admitted to hospital before and 1 year after the publication of national guidelines. Respir Med. 1996;90:539–545. Abstract |
Full-Text PDF (629 KB)
|
CrossRef
3. 3Guevara JP, Wolf FM, Grum CM, Clark NM. Effects of educational interventions for self management of asthma in children and adolescents: systematic review and meta-analysis. BMJ. 2003;326:1308–1309. 4. 4Salmeron S, Liard R, Elkharrat D, Muir J, Neukirch F, Ellrodt A. Asthma severity and adequacy of management in accident and emergency departments in France: a prospective study. Lancet. 2001;358:629–635. Abstract | Full Text |
Full-Text PDF (288 KB)
|
CrossRef
5. 5Eccles M, McColl E, Steen N, Rousseau N, Grimshaw J, Parkin D, et al. Effect of computerised evidence based guidelines on management of asthma and angina in adults in primary care: cluster randomised controlled trial. BMJ. 2002;325:941. 6. 6Sithamparanathan M. That’s all I got handed over. BMJ. 2006;332:496. 7. 7Grol R, Grimshaw J. From best evidence to best practice: effective implementation of change in patients’ care. Lancet. 2003;362:1225–1230. Abstract | Full Text |
Full-Text PDF (106 KB)
|
CrossRef
8. 8Campbell H, Hotchkiss R, Bradshaw N, Porteous M. Integrated care pathways. BMJ. 1998;316:133–137. 9. 9Sulch D, Perez I, Melbourn A, Kalra L. Randomized controlled trial of integrated (managed) care pathway for stroke rehabilitation. Stroke. 2000;31:1929–1934. MEDLINE 10. 10Kwan-Gett TS, Lozano P, Mullin K, Marcuse EK. One-year experience with an inpatient asthma clinical pathway. Arch Pediatr Adolesc Med. 1997;151:684–689. MEDLINE 11. 11McDowell KM, Chatburn RL, Myers TR, O’Riordan MA, Kercsmar CM. A cost-saving algorithm for children hospitalized for status asthmaticus. Arch Pediatr Adolesc Med. 1998;152:977–984. MEDLINE 12. 12Kelly CS, Andersen CL, Pestian JP, Wenger AD, Finch AB, Strope GL, et al. Improved outcomes for hospitalized asthmatic children using a clinical pathway. Ann Allergy Asthma Immunol. 2000;84:509–516. Abstract |
Full-Text PDF (190 KB)
|
CrossRef
13. 13Wazeka A, Valacer DJ, Cooper M, Caplan DW, DiMaio M. Impact of a pediatric asthma clinical pathway on hospital cost and length of stay. Pediatr Pulmonol. 2001;32:211–216. MEDLINE |
CrossRef
14. 14Norton SP, Pusic MV, Taha F, Heathcote S, Carleton BC. Effect of a clinical pathway on the hospitalization rates of children with asthma: a prospective study. Arch Dis Child. 2007;92:60–66.
CrossRef
15. 15Johnson KB, Blaisdell CJ, Walker A, Eggleston P. Effectiveness of a clinical pathway for inpatient asthma management. Pediatrics. 2000;106:1006–1012. 16. 16British guideline on the management of asthma. Thorax. 2003;58(Suppl 1):i1–i94. 17. 17Dodds C. Assessing pharmacy interventions at Salisbury Healthcare NHS Trust. Hosp Pharmacist. 2003;19:451–456. 18. 18Brown H, Prescott R. Applied mixed models in medicine. Chichester, UK: John Wiley & Sons; 2006;. 19. 19Berwick DM. Disseminating innovations in health care. JAMA. 2003;289:1969–1975. MEDLINE |
CrossRef
20. 20Teach SJ, Crain EF, Quint DM, Hylan ML, Joseph JG. Improved asthma outcomes in a high-morbidity pediatric population: results of an emergency department-based randomized clinical trial. Arch Pediatr Adolesc Med. 2006;160:535–541. MEDLINE |
CrossRef
21. 21Mitchell EA, Didsbury PB, Kruithof N, Robinson E, Milmine M, Barry M, et al. A randomized controlled trial of an asthma clinical pathway for children in general practice. Acta Paediatr. 2005;94:226–233. MEDLINE |
CrossRef
22. 22Wesseldine LJ, McCarthy P, Silverman M. Structured discharge procedure for children admitted to hospital with acute asthma: a randomized controlled trial of nursing practice. Arch Dis Child. 1999;80:110–114.
CrossRef
23. 23Madge P, McColl J, Paton J. Impact of a nurse-led home management training program in children admitted to hospital with acute asthma: a randomized controlled study. Thorax. 1997;52:223–228. MEDLINE |
CrossRef
24. 24Dean B, Schachter M, Vincent C, Barber N. Causes of prescribing errors in hospital inpatients: a prospective study. Lancet. 2002;359:1373–1378. Abstract | Full Text |
Full-Text PDF (92 KB)
|
CrossRef
a Department of Respiratory and Sleep Medicine, Royal Hospital for Sick Children, Edinburgh, United Kingdom b Medical Statistics Unit, University of Edinburgh, Edinburgh, United Kingdom c Department of Emergency Medicine, Royal Hospital for Sick Children, Edinburgh, United Kingdom.  Reprint requests: Dr Steve Cunningham, Consultant Respiratory Paediatrician and Part Time Senior Lecturer, Department of Respiratory and Sleep Medicine, Royal Hospital for Sick Children, Sciennes Road, Edinburgh, EH9 1LF, UK.
PII: S0022-3476(07)00893-1 doi:10.1016/j.jpeds.2007.09.033 © 2008 Mosby, Inc. All rights reserved. | |
|
FULL TEXT
