POSSUM scoring for patients with fractured neck of femur{dagger}

T. S. Ramanathan1, I. K. Moppett1,*, R. Wenn2 and C. G. Moran2

1 University Department of Anaesthesia and 2 Trauma and Orthopaedics Directorate, Queen's Medical Centre, Nottingham NG7 2UH, UK

* Corresponding author. E-mail: Iain.Moppett{at}nottingham.ac.uk

Accepted for publication December 2, 2004.


    Abstract
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Background. POSSUM scoring is validated as an audit tool in general and orthopaedic surgery. It is also used for preoperative triage to assess perioperative risk. However its ability to predict mortality in specific surgical subgroups, such as patients with fractured neck of the femur, has not been studied. This study assessed the predictive capability of POSSUM for 30-day mortality after surgery for fractured neck of femur.

Methods. A cohort study was conducted in Queen's Medical Centre, Nottingham over a period of nearly 2 yr. Complete data from 1164 patients were analysed to compare the mortality predicted by POSSUM and the observed mortality. POSSUM risk of death was calculated using the original POSSUM equation, with modifications to the operative score appropriate for orthopaedic surgery.

Results. POSSUM predicted 181 (15.6%) deaths and the observed mortality was 119 (10.2%). The area under the receiver operating characteristic curve was 0.62, indicating poor performance by the POSSUM equation.

Conclusion. POSSUM overpredicts mortality in hip fracture patients. It should be used with caution whether as an audit tool or for preoperative triage.

Keywords: audit, POSSUM scoring ; complications, fractured neck of femur ; complications, mortality


    Introduction
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
The Physiological and Operative Severity Score for enUmeration of Mortality and morbidity (POSSUM) was originally designed by Copeland and colleagues to assess outcome after general surgery.1 Twelve physiological and six operative factors are scored, and are subjected to logistic regression analysis to generate two equations to predict population mortality and morbidity. Since the original study, various modifications to the score have been made which may provide better prediction in specific groups of patients.25 The original POSSUM equation (using a modified operation classification) has been validated in orthopaedic surgery.5

In the current health-care climate, there is increasing use of scoring systems. First, they are used to compare operative outcome both between and within units. Secondly, scoring systems are used to identify high-risk patients before surgery, to inform the consent process, and to triage the use of higher-level care. This preoperative use of POSSUM appears anecdotally to be increasing despite being outside the original validation of the system. Patients presenting with fractured neck of the femur form a large, high-risk group for which POSSUM would seem useful. However, POSSUM has not been formally evaluated for postoperative or preoperative use in fractured neck of femur surgery.


    Methods
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
From August 2001 to May 2003, all patients admitted with fractured neck of femur to Queen's Medical Centre have had prospective collection of physiological and operative data, with the intention of auditing factors associated with morbidity and mortality in this population. Data were collected manually by dedicated audit officers from hospital computer and paper records. Blood loss measurement was recorded routinely at the time of operation. The surgical drapes used have an integral blood collection bag, and all soiled swabs are weighed. Thirty-day mortality data were collected and cross-checked from hospital statistics and the Office of National Statistics. Morbidity data collected do not cover the complete POSSUM set so were not examined in this study.

All data were entered into a Microsoft Excel XP® spreadsheet and POSSUM scores generated using automated calculations. A random sample of data (20 records) was checked manually by one of the investigators (I.K.M.), as were the highest 5% and lowest 5% of POSSUM scores, both for accuracy of input data and correct calculation of physiological, operative and total scores. Individual physiological values which deviated markedly from normal were cross-checked with hospital records. No errors were found. Previous internal cross-checking of this audit data has found an error rate of <3%.

Each physiological datum is given a score between 1 and 8 using the original POSSUM system (Table 1),1 giving a minimum and maximum physiological score of 12 and 96. The operative data are scored using a modification of the original POSSUM system to allow for orthopaedic operations (Table 2),2 again between 1 and 8, giving a minimum of 6 and maximum of 48. Routine surgery for fractured neck of femur (dynamic hip screw, hemiarthroplasty) is scored as ‘major’ (4 points); revision surgery is ‘major+’ (8 points). The total physiological and operative scores are then entered into a logistic regression equation, which gives a risk of death.


View this table:
[in this window]
[in a new window]
 
Table 1 POSSUM physiology score. Points are given for each datum according to the table. JVP, jugular venous pressure; SOB, shortness of breath; COAD, chronic obstructive airways disease; WBC, white blood count

 

View this table:
[in this window]
[in a new window]
 
Table 2 POSSUM operative severity score adjusted for orthopaedic surgery

 
Risk of death and observed death rates were compared for increasing increments of risk. A standard receiver operating characteristic curve (ROC) was generated, plotting sensitivity vs 1–specificity. Using previous estimates of mortality of approximately 10%, we estimated that a total sample size of 1500 subjects would give a power of 0.8 with alpha error of 0.05 to detect an area under the curve (AUC) of 0.65. Goodness-of-fit testing was undertaken using the method of Lemeshow and Hosmer.6 Statistical analysis was performed using Minitab v13 (Minitab, Pennsylvania, USA).


    Results
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
Over the 21 months of the study, 1315 patients were admitted to Queen's Medical Centre with fractured neck of the femur. Most (86%) of the patients were aged over 70 yr and 75% were female. Those admitted from their own home had a lower mortality than those admitted from nursing home or rest home ({chi}2 test, P<0.001) Complete data were available for 1164 patients. There were 151 patients excluded from the study; 56 were treated conservatively, of whom 50% died within 30 days, and all had died by 15 months. In those who underwent operation, 76 had missing data, 16 were operated on 48 h after admission, and three were lost to follow-up due to transfer to other hospitals. There is no scoring suggested in the orthopaedic POSSUM for those operated on after 48 h. These patients fall into a clinically separate group as they are almost always either current hospital in-patients, transferred over to orthopaedic care after a more delayed diagnosis, or have had surgery postponed for medical reasons. We therefore felt it would be inappropriate to include them. Only four patients had operations within 6 h, all of whom were young (19–52 yr). Those with incomplete data did not have their POSSUM risk of death calculated, but physiological scores were calculated where possible. POSSUM overpredicted death overall [181 (POSSUM) vs 119 (observed)]. Comparison between POSSUM prediction and observed mortality for increasing levels of risk is given in Table 3. Goodness-of-fit testing using the method of Lemeshow and Hosmer6 indicated a poor fit ({chi}2 test, P<0.00015). Using predicted 50% risk of death, specificity was 98.3%, sensitivity 7.6%, positive predictive value 33.3% and negative predictive value 90.3%.


View this table:
[in this window]
[in a new window]
 
Table 3 Distribution of patients based on the POSSUM predicted risk band

 
The ROC curve is shown in Figure 1. The AUC was 0.62 (SE 0.024), indicating poor predictive value for the POSSUM score.



View larger version (10K):
[in this window]
[in a new window]
 
Fig 1 Receiver operating characteristic curve for 30-day mortality after surgery for fractured neck of femur. The line of identity, which represents a non-discriminatory test, is shown with a dashed line. A perfect test would follow a right-angle with an area under the curve of 1. The area under the curve for POSSUM is 0.62.

 

    Discussion
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
POSSUM was designed originally as a postoperative, general surgical, audit tool. However, its use has since expanded into other surgical fields and preoperative assessment. In theory, POSSUM provides an attractive tool for surgical audit and triage. However, if it is to be used with confidence, POSSUM needs to be validated for specific surgical procedures.

Overall, we found that POSSUM overestimated the risk of death, particularly in those patients with a higher predicted risk of dying. The ROC area under the curve of 0.62 indicates a poor test, particularly in the context of AUC values for POSSUM in general surgery of >0.97 and general orthopaedics of >0.85.5 Other workers have found a similar overprediction of POSSUM when used in both general surgery and specific operative subgroups.812 Our results are based on a single centre, so there is the possibility that our results are at variance with national/international experience. This cannot be excluded; however, the 30-day mortality in this series of 10.2% is in line with national statistics and other published data.13

In the original POSSUM, the physiological data were collected close to the surgery, whereas we collected the data on hospital admission. Given the relatively short period from admission to operation, this is unlikely to have a marked impact on results. One might expect an improvement in physiological score following admission and/or resuscitation, which could explain some of the overprediction from POSSUM. Conversely, patients may deteriorate after admission, which would lead to underestimation of physiological dysfunction. These issues apply to all scoring systems and to date there appears to be no consensus on the best time to estimate risk.14 Clearly, the presence of missing data may also have altered the results. However, including all patients operated on does not change the overall 30-day mortality significantly (10.5 vs 10.2%). High-risk patients may also have been identified early and provided with appropriate high-level perioperative anaesthetic, orthopaedic and nursing care, hence reducing their mortality. This is unlikely, given the similarity of overall mortality between this and other series. Also in our centre, all patients with fractured neck of femur have surgery on dedicated trauma theatres, performed by anaesthetists and surgeons of specialist registrar grade and above. Because of the homogeneity of surgery in this group of patients, the surgical component of POSSUM has a very narrow range, so the variability in total POSSUM is largely due to changes in physiological (preoperative) score. Some authors have found that using purely the physiological component of POSSUM performs as well as the combined scores in predicting mortality in ruptured abdominal aortic aneurysm4 and major arterial surgery.3 4 In our series, this performs poorly. There are probably several reasons for this.

The patient population is generally elderly. Fifty-nine per cent of them were >80 yr. Hence, the presence of abnormalities on blood tests or examination may be more ‘normal’ than in a general, unselected population and hence not a specific marker of at-risk patients. The weighting of individual components may also be inappropriate for this population, with undue emphasis on certain aspects.14 Previously published work has found strong associations between serum albumin or haemoglobin concentrations on admission and mortality in this group.15 16 Whatever the reason, POSSUM is not identifying the correct risk factors for this patient group.

Recently published work has suggested the use of a simplified scoring system, which performs better than POSSUM in a general surgical population.17 We did not formally test this new score with our data. However, for this relatively homogeneous population, it is likely to perform badly. The score uses four factors: age, ASA grading, mode of surgery (elective vs emergency) and severity (three-point scale). All surgery for fractured neck of the femur is emergency surgery, and all the operations are classed as Grade II. Most patients are ASA II or III and most are >70 yr of age. Thus, there is very little variation possible within the Donati score.

Does it matter whether POSSUM is valid for use in this population? Surgical and anaesthetic audit is an appropriate part of modern medical practice. Good audit requires some degree of gold standard against which to compare results. The temptation is to extrapolate validated audit tools into other fields. We would suggest that POSSUM does not reliably perform this role. Using POSSUM to identify high-risk patients before surgery is also not reliable. POSSUM overpredicts mortality in fractured neck of the femur, and should not be used as a comparative audit tool for this group of patients. Its role as a preoperative assessment tool is also limited. Further work combining established risk factors for this population, such as concentrations of haemoglobin16 and serum albumin15 may provide a better predictive tool.


    Footnotes
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
{dagger} An abstract of part of the study was presented at the Anaesthetic Research Society meeting, Aberdeen, April 2004 and published in British Journal of Anaesthesia 2004; 93: 161. Back


    References
 Top
 Footnotes
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Copeland GP, Jones D, Walters M. POSSUM: a scoring system for surgical audit. Br J Surg 1991; 78: 355–60[Medline]

2 Prytherch DR, Whiteley MS, Higgins B, Weaver PC, Prout WG, Powell SJ. POSSUM and Portsmouth POSSUM for predicting mortality. Br J Surg 1998; 85: 1217–20[CrossRef][ISI][Medline]

3 Prytherch DR, Ridler BM, Beard JD, Earnshaw JJ. A model for national outcome audit in vascular surgery. Eur J Vasc Endovasc Surg 2001; 21: 477–83[CrossRef][ISI][Medline]

4 Prytherch DR, Sutton GL, Boyle JR. Portsmouth POSSUM models for abdominal aortic aneurysm surgery. Br J Surg 2001; 8: 58–63

5 Mohamed K, Copeland GP, Boot DA, et al. An assessment of the POSSUM system in orthopaedic surgery. J Bone Joint Surg (Br) 2002; 84B: 35–9

6 Lemeshow S, Hosmer DW Jr. A review of goodness of fit statistics for use in the development of logistic regression models. Am J Epidemiol 1982; 115: 92–106[Abstract]

7 Mohil RS, Bhatnagar D, Bahadur L, Rajneesh, Dev DK, Magan M. POSSUM and P-POSSUM for risk-adjusted audit of patients undergoing emergency laparotomy. Br J Surg 2004; 91: 00–3

8 Lam CM, Fan ST, Yuen AW, Law WL, Poon K. Validation of POSSUM scoring systems for audit of major hepatectomy. Br J Surg 2004: 91: 50–4

9 Khan AW, Shah SR, Agarwal AK, Davidson BR. Evaluation of the POSSUM scoring system for comparative audit in pancreatic surgery. Digestive Surgery 2003; 20: 539–45[CrossRef][ISI][Medline]

10 Senagore AJ, Delaney CP, Duepree HJ, Brady KM, Fazio VW. Evaluation of POSSUM and P-POSSUM scoring systems in assessing outcome after laparoscopic colectomy. Br J Surg 2003; 90: 1280–4[CrossRef][ISI][Medline]

11 Tambyraja AL, Kumar S, Nixon SJ. The Physiological and Operative Severity Score for the enUmeration of Mortality and morbidity (POSSUM). Br J Surg 2003; 90: 157–65[CrossRef][ISI][Medline]

12 Yii MK, Ng KJ. Risk-adjusted surgical audit with the POSSUM scoring system in a developing country. Physiological and Operative Severity Score for the enUmeration of Mortality and morbidity. Br J Surg 2002; 89: 110–3[CrossRef][ISI][Medline]

13 Roberts SE, Goldacre MJ. Time trends and demography of mortality after fractured neck of femur in an English population, 1968–98: database study. Br Med J 2003; 327: 771–5[Abstract/Free Full Text]

14 Neary WD, Heather BP, Earnshaw JJ. The Physiological and Operative Severity Score for the enUmeration of Mortality and morbidity (POSSUM). Br J Surg 2003; 90: 157–65[CrossRef][ISI][Medline]

15 Koval KJ, Maurer SG, Edward T, et al. The effects of nutritional status on outcome after hip fracture. J Orthop Trauma 1999; 13: 164–9[CrossRef][ISI][Medline]

16 Gruson KI, Aharonoff GB, Egol KA, et al. The relationship between admission haemoglobin level and outcome after hip fracture. J Orthop Trauma 2002; 16: 39–44[CrossRef][ISI][Medline]

17 Donati A, Ruzzi M, Adrario E, et al. A new and feasible model for predicting operative risk. Br J Anaesth 2004; 93: 393–9[Abstract/Free Full Text]





This Article
Abstract
Full Text (PDF)
All Versions of this Article:
94/4/430    most recent
aei064v1
E-Letters: Submit a response to the article
Alert me when this article is cited
Alert me when E-letters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Disclaimer
Request Permissions
Google Scholar
Articles by Ramanathan, T. S.
Articles by Moran, C. G.
PubMed
PubMed Citation
Articles by Ramanathan, T. S.
Articles by Moran, C. G.