Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
Delay in surgery alone doesn’t seem to increase mortality after hip fracture, with adjustment for comorbidity, in Friuli Venezia Giulia (Italy)
- Antonella Franzo, Giorgio Simon (3 April 2006)
-
Delay to hip fracture surgery – Scottish data provide additional information
- D Graham Mackenzie, Sarah Wild and Rod Muir (25 April 2006)
-
We have had the best year ever...
- Stephen P Hodgson, Simon Stacey, Consultant Physician and Geriatrician (2 May 2006)
-
Defining delay
- SEEMA QUASIM (8 May 2006)
-
Further evidence from routine data by using the hospital-level hypotheses
- Reijo Sund (7 August 2006)
|
|
|||
|
Antonella Franzo, Epidemiologist Agenzia Regionale della Sanità, piazzale S. Maria della Misericordia 15 33100 Udine Italy, Giorgio Simon
Send response to journal:
|
We estimated the association between delay in surgery and in-hospital mortality in Friuli Venezia Giulia, a region located in the North-East of Italy with approximately 1,200,000 resident population, where each year more than 2,200 elderly people are admitted to regional hospitals for hip fracture.
We examined data from regional hospital discharge register and all patients aged 65 and over admitted to regional hospitals between 1996 and 2005 with a main diagnosis of hip fracture and surgically treated were selected. Patients with a malignant neoplasm diagnosis were excluded. In case of multiple hip fracture episodes only the first one was included. The total number of patients included was 13,822. To assess comorbidity, all diagnoses coded in the hospital discharge records and in those of the previous year were considered. Based on these data, the Charlson comorbidity index was calculated for each patient (1-2) and transformed in a dichotomic variable, with 1 indicating presence of comorbidity and 0 absence of comorbidity The waiting time for surgery was dichotomized following the clinical indication of the medical literature (1 indicated immediate surgical treatment (same day of hospital admission or 1 day after admission) and the value 2 indicated delayed surgical treatment (2 days and more after hospital admission). The year in which the fracture occurred was included in the multivariate models in order to evaluate changing trends in mortality, after controlling for case mix of patients and modeled as a continuous variable. We estimated non conditional logistic regression models as well as several multilevel logistical models to take into account hospital level variability (3-5). We used a logistic regression model with robust variance estimates (Hubner/White/sandwich estimator) specifying hospitals as a cluster variable in order to obtain confidence intervals for the odd ratio adjusted for intraclass correlation. Confidence intervals were provided. The analysis of data was performed using Stata 7.0. The cohort selected was composed of 13,822 patients. A proportion of 4.7% of patients died during hospital stay. Bivariate analysis showed a significant association between mortality and age, male sex, comorbidity and delayed surgery and a decrease in mortality from 1996 to 2005. Multilevel models, taking into account hospital level variability, showed no significant association between delayed surgery and increasing in-hospital mortality (OR 1,18 CI 95%0.84-1.65). Comorbidity, male sex and advanced age were associated with an increased in-hospital and post-discharge mortality in all estimated models. Between 1996 and 2005 mortality decreased significantly. These results do not threaten the validity of clinical guidelines that recommend a surgical treatment for hip fracture within 24 hours after hospital admission. In fact, immediate surgery can reduce immobilization time, the risk of pressure sores, infections and thromboembolic complications. However, some delay in surgery may be due to the severity of the health conditions of patients, which require adequate therapies and complex clinical inquiries. Delay related to hospital organizational problems should be avoided. The substantial reduction of mortality risk after hip fracture surgery between 1996 and 2005 may be due to improvements in surgical, and anesthesiological techniques or to an extended application of thromboembolism, and hospital infection prevention protocols. From a methodological point of view, our study is characterized by a population-based approach, and a complete cohort of residents in Friuli Venezia-Giulia region was selected. The total amount of patients analyzed (13,822) is quite large and although the data come from administrative database, we believe that case mix measures are reliable and sufficiently detailed. In conclusion, our data suggest that delayed surgery alone doesn’t seem to increase mortality after hip fracture, with adjustment for comorbidity, age and sex and taking into account hospital level variability. References 1. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40(5):373-83. 2. Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 1992;45(6):613-9. 3. Hosmer DH, Lemeshow S. Applied logistic regression. New York: John Wiley & Sons, 1989. 4. Diggle PJ, Zeger SL, Liang KY. Analysis of longitudinal data. Oxford (USA): Clarendon Press, 1994. 5. Diez Roux AV. A glossary for multilevel analysis. J Epidemiol Community Health 2002; 56: 588-594. Competing interests: None declared |
|||
|
|
|||
|
D Graham Mackenzie, Specialist Registrar in Public Health Public Health Department, NHS Fife, Cameron House, Cameron Bridge, Windygates, Leven, KY8 5RG, Sarah Wild and Rod Muir
Send response to journal:
|
Editor - Bottle and Aylin’s study provides evidence of an association between delay to hip fracture surgery and mortality1, a topic that has proved controversial2. However, as Bottle and Aylin’s analysis depended on routinely collected data, there were important limitations to the study – namely, the lack of information about the reason for delay to surgery and the short period of follow up. Scottish data provide additional relevant information.
Scottish Hip Fracture Audit data are collected prospectively3, and records have been linked to routinely collected data for hospital admissions and mortality4. We have studied one-year mortality in patients undergoing hip fracture surgery between 1998 and 2003, in groups of patients stratified by the reason for delay to surgery (n=8470). We used Cox proportional hazards regression analysis to adjust for potential confounding factors including age, sex, hospital, pre-fracture residence, American Society of Anesthesiologists grade (a measure of systemic illness prior to fracture), and number and type of hospital admissions in the five years prior to hip fracture. Surgery was delayed by more than 24 hours after hospital admission in 3364 (40%) patients of which 1432 operations (43%) were delayed for administrative reasons including restricted access to theatre, surgeon or anaesthetist, 1315 people (39%) were initially medically unfit for surgery and 617 operations (18%) were delayed for other reasons. Overall, 2531 (30%) patients died in the year following hip fracture admission. Compared to patients operated on within 24 hours, delay to surgery in patients who were initially medically unfit was associated with increased mortality (hazard ratio 1.3; 95% confidence interval 1.1 to 1.4). However, there was no evidence of an association between delay to surgery and mortality for patients whose operation was delayed for administrative reasons (HR 0.9, 95% CI 0.8 to 1.0) or for other reasons (HR 1.1, 95% CI 0.9 to 1.2). These Scottish data support an association between delay to surgery and mortality following hip fracture, and show that the association persisted for at least a year following hip fracture admission. However the excess mortality appears only to be present when the delay was for medical reasons. The importance of delay to hip fracture surgery therefore remains open to debate. D Graham Mackenzie, Specialist Registrar in Public Health, Public Health Department, NHS Fife, Cameron House, Cameron Bridge, Windygates, Leven, KY8 5RG Sarah Wild, Senior Lecturer in Public Health and Epidemiology, School of Clinical Sciences and Community Health, University of Edinburgh, Edinburgh, EH8 9AG Rod Muir, Consultant in Public Health Medicine, Information Services, NHS National Services Scotland, Gyle Square, 1 South Gyle Crescent, Edinburgh, EH12 9EB References
Competing interests: None declared |
|||
|
|
|||
|
Stephen P Hodgson, Consultant Orthopaedic Surgeon Royal Bolton Hospital BL4 0JR, Simon Stacey, Consultant Physician and Geriatrician
Send response to journal:
|
RE; Mortality associated with delay in operation after hip fractures We feel that this study demonstrates significant potential for improvement in the management of hip fractures in England. The data reveals that even excluding patients from residential care the 30 day mortality rate is 14.9% and that 40% of procedures were performed more than one day and 22% more than two days after admission. There are large variations between hospitals with regard to delay(16% -51.2% having surgery within two days of admission)1. At the Royal Bolton Hospital there had been concern regarding our hip fracture service with a high incidence of delays from admission to operation and a high 30 day mortality rate. This resulted in intense focus on hip fracture care across trust management, orthopaedics and elderly medicine. As a result service improvements have been introduced including additional trauma lists and the introduction of a trauma stabilisation unit (orthopaedic high dependency unit) with daily consultant orthogeriatric input. These measures have reduced average delay from admission to operation from 2.3 days to 1.7 days. There has also been improvement in our 30 day mortality figures (RR of death 1.74 compared to peer in the year 2004-2005 vs. RR 1.05 for the year 2005-2006) which we hope will be sustained. We do feel confident that we have improved the quality of service for this vulnerable group of patients and that further improvement will be possible. 1. Alex Bottle and Paul Aylin .Mortality associated with delay in operation after hip fracture: observational study BMJ 2006; 332: 947-951 Mr Stephen Hodgson
Dr. Simon Stacey
Competing interests: None declared |
|||
|
|
|||
|
SEEMA QUASIM, SPR Anaesthetics Good Hope Hospital, Sutton Coldfield, B75 7RR
Send response to journal:
|
I was interested to read this study but must comment on a methodological flaw. Hospital statistic data was used to determine delay but the only the date of admission and operation was taken into account. This is misleading. For example, if admission occurs today and operation occurs tomorrow, the delay would be apparently one day. But if the patient had been admitted at 0100h and operated at 2100h the following day, the operative delay is actually 32 hours (i.e. more than one day if one defines a day as 24 hours). I recently encountered this problem in a similar audit on fracture neck of femur patients, albeit at local level. Initially, a delay (in terms of days) did appear to have an effect on mortality, but once the data was re-analysed using times as well as dates, this effect completely disappeared and the original conclusion proved wrong. Therefore I do not believe that this study adds anything to the debate over whether a delay in surgery in these patients does adversely affect mortality. Competing interests: None declared |
|||
|
|
|||
|
Reijo Sund, Statistician National Research and Development Centre for Welfare and Health, Helsinki, Finland
Send response to journal:
|
Bottle and Aylin found significant association between delay to hip fracture operation and mortality in individual based analyses, but they report that there were no significant relations between highly varying delay rates and the proportion of mortality at the hospital trust level (1). However, a simple performance assessment interpretation of hospital specific delay rates is that the percentage of late surgery patients can be reduced to a potentially achievable level (2) that can also be interpreted as the upper limit for the proportion of acceptable delayed patients. Correspondingly, the expected proportion of unacceptable delayed patients is the proportion of late surgery patients exceeding this upper limit, and thus the overall hospital level mortality of hip fracture patients should increase with an increasing proportion of late surgery patients given that the longer operative delay would have an adverse effect on mortality. In this sense, the results of individual level and hospital level analyses of Bottle and Aylin are somewhat conflicting. Another hospital level hypothesis is that the mortality of late surgery patients is higher if only the patients unfit for surgery are delayed, since the unfit condition for surgery is also a risk factor for mortality. The analysis of 16 881 operated hip fracture patients from 47 hospitals in Finland in 1998-2001 revealed that a smaller proportion of late surgery patients was non-linearly associated with a higher mortality rate for these patients (3). This is an indication that the effect of the operative delay on mortality is mainly due to the unavoidable delay for medical reasons. In conclusion, the results based on individual level routine data incorporating several providers are prone to serious bias, but the use of improved methodology including hospital level hypotheses and analyses helps to overcome certain limitations of the routine data (3). Reijo Sund, Statistician, National Research and Development Centre for Welfare and Health, Helsinki, Finland References 1. Bottle A, Aylin P. Mortality associated with delay in operation after hip fracture: observational study. Bmj 2006;332(7547):947-51. 2. Gibberd R, Hancock S, Howley P, Richards K. Using indicators to quantify the potential to improve the quality of health care. Int J Qual Health Care 2004;16(Suppl 1):i37-43. 3. Sund R, Liski A. Quality effects of operative delay on mortality in hip fracture treatment. Qual Saf Health Care 2005;14(5):371-7. Competing interests: None declared |
|||