Are Ottawa Knee Rules Useful in Actual Trauma Care?

DOI: 10.2478/s11536-007-0013-7 Research article CEJMed 2(2) 2007 216–221

Are Ottawa Knee rules useful in actual trauma care?

Kamal Nagpal∗, Mandar Marathe Doncaster Royal Inﬁrmary, Thorne Road, Doncaster DN2 5LB, Doncaster, UK

Received 3 December 2006; accepted 28 January 2007

Abstract: Acute knee injuries are a common presentation to the emergency department(ED). Ottawa knee rules (OKR) have shown to reduce the number of radiographs in these patients in North American studies and a fracture rate of 5% has been reported. Based on this, we tested the hypothesis that it was possible to decrease the number of x-ray ﬁlms obtained after a knee trauma without delayed fracture diagnosis by means of the Ottawa knee rules in British set up. A total of 118 adult patients with acute Knee injuries were studied. A checklist in an easy-to-use format was produced to act as an aide-memoir and to encourage clinicians to apply the OKR in their decision making. Sixty patients were studied before introducing the check list stickers of OKR and ﬁfty eight were studied after introducing the stickers. The OKR were found to have been obeyed in 24 (40%) of patients in the group which did not have stickers. In the group assessed without an OKR sticker, 28 (46.7%) of patients had knee radiography, compared with 29 (50%) of patients in the group who were assessed with an OKR sticker. There was no decrease in the number of x-rays after following the Ottawa Knee Rules (OKR). c Versita Warsaw and Springer-Verlag Berlin Heidelberg. All rights reserved.

Keywords: Ottawa Knee rules, acute knee injury, knee rules

1 Introduction

Clinical decision-making rules are tools designed to assist physicians. They have the definite potential to significantly reduce health care costs while maintaining the optimum level of patient care. In trauma care, it has been demonstrated that the use of clinical rules was able to decrease the need for radiography after ankle trauma without significant inconvenience [1]. Acute knee trauma is very common. It is accepted to perform routine plain radiography after such a trauma to detect any such fracture and for possible medico

∗ E-mail: kamnag76@yahoo. com K. Nagpal, M. Marathe / Central European Journal of Medicine 2(2) 2007 216–221 217 legal issues. However, most of these x-ray studies yield normal findings and bring little additional information. A fracture rate of about 5% in acute knee injury patients has been reported [2]. Specific imaging like an MRI scan is necessary to rule out ligamental or meniscal injuries. One can therefore wonder if routine radiography after knee trauma is actually cost-effective. The Ottawa knee rules (OKR) are a set of clinical decision rules for radiography in acute knee injury [3](Figure1). They aim to help clinicians to be more selective when requesting knee radiographs in two planes in adults. North American studies have shown the rules to be both valid [3, 4] and to reduce the number of unnecessary radiographs performed [5, 6]. It is not clear whether this decision rule would be useful in British practice. We wanted to test the hypothesis that it was possible to use these rules in our district general hospital (Level 2 trauma centre) to decrease the number of performed x-rays studies without missing any fracture diagnosis.

Fig. 1

2 Statistical methods and Experimental Procedures

This was a observational study in the ED of a district general hospital in England over 4- month period. There were two phases: phase1 before and phase 2 after the introduction of the OKR. A total of 118 patients with acute knee injuries were studied. Inclusion criteria were patients with acute knee injuries due to blunt trauma. Patients with lacerations, contusions only and due to penetrating injury were excluded from the study. Children were also excluded from the study. Sixty patients were studied before introducing the check list of OKR and fifty eight were studied after introducing the check list The ED case notes from consecutive ED attendances over an eight weeks period were obtained, where the principle diagnosis on discharge, as recorded by the clinician, was either knee sprain or knee fracture. Case notes were studied and information was recorded on a standard proforma in- cluding age, gender, grade of clinician. It also included whether a knee radiograph was obtained, whether the documented clinical examination findings supported the decision to perform knee radiography or not perform knee radiography according to the OKR, and whether a knee fracture was detected on the radiograph (confirmed by the radiologist report). 218 K. Nagpal, M. Marathe / Central European Journal of Medicine 2(2) 2007 216–221

Using a personal computer running the Microsoft Windows operating system and Microsoft Word word-processing software, a checklist was produced to act as an aide- memoir and to encourage clinicians to apply the OKR in their decision making (see Figure 1). This checklist highlights the OKR in an easy-to- use format. The checklist was printed with a HP LaserJet office printer onto self-adhesive labels (Avery product code L7165, 8 labels per A4 sheet, each label 99.1 mm × 67.7 mm). A supply of labels was placed prominently in the triage room of the ED. Triage nurses were informally instructed in the use of the OKR and the stickers. Specifically, they were instructed to place a sticker on the case notes of each patient presenting with a knee injury, and who met the inclusion criteria. Junior, senior ED doctors and Emergency nurse practitioners were also informed on the use of the sticker checklist and were made aware of and usage of OKR. Case notes from ED attendances over the subsequent eight-week period were obtained and assessed using the same criteria as described previously, where the ED case notes contained an OKR sticker-checklist. Data were analysed using SPSS SigmaStat for Windows Version 3. 0 statistical analysis software to examine for differences between the group of patients who were assessed without an OKR sticker and the group of patients assessed with an OKR sticker.

3Results

During the study period of March 2005 to June 2005, all 118 eligible patients with knee injuries were included in the study. 49% were males and rest were females. Of the 118 patients, 7 (5.9%) had a final diagnosis of bony injury. In phase 1 of the trial, without OKR sticker, 60 patients were eligible for the study on attendance at the ED. In phase2, after the introduction of OKR, 58 patients were eligible for the study. There was no significant difference in age and gender distribution, grade of clinician between the patients who were assessed without an OKR sticker and those who were assessed with the help of an OKR sticker. The OKR were found to have been obeyed in 24 (40%) of patients in the group which did not have stickers, compared with 100% of patients in the group which did have stickers Chi-square (with Yates’ correction for continuity) = 47.288 with 1 degrees of freedom. P <0.001. These findings are summarised in Table 1.

Table 1 Contingency table showing whether OKR were followed.

No sticker Sticker Total

OKR followed 24 / 60 (40%) 58/58 (100%) 82/118 OKR not followed 36/60(60%) 0/58(0%) 36 / 118 Total 60/118 58/118 K. Nagpal, M. Marathe / Central European Journal of Medicine 2(2) 2007 216–221 219

In the group assessed without an OKR sticker, 28 (46.7%) of patients had knee radiography, compared with 29 (50%) of patients in the group who were assessed with an OKR sticker having knee radiography {Chi-square (with Yates’ correction for continuity) = 0.032 with 1 degrees of freedom, P = 0.859}. In the group of patients assessed without an OKR sticker, 4 patients (6.7%) had an knee fracture, compared with 3 patients (5.2%) in the group assessed with an OKR sticker*** {Chi-square (with Yates’ correction for continuity) = 0.002 with 1 degrees of freedom. P = 0 963}.(Table2)

Table 2 Radiography requesting pattern and Outcome.

Without OKR With OKR Total Sticker Sticker

Radiograph 28 29 57 No radiograph 32 29 61 Bony injury 437 No bony injury 56 55 111

4 Discussion

The index study of the Ottawa rules for knee trauma [4] involved a series of 1096 cases. The authors observed the following results: sensitivity of 100%, specificity of 49%, positive predicitive value of 11%, and negative predictive value of 100%. The observed reduction in use of radiography was estimated to be 28%. The rules were introduced and previously validated in north America, where other studies have shown the sensitivity to be between 84. 6% [7], 97% [8], 98% [9], and 100% [3]. This study shows the rules to be an effective tool in deciding which patients with isolated acute knee injury do not require knee radiography when used by doctors in ED. Previous studies have shown a reduction in the number of radiographs requested after the introduction of the OKR [4, 6]. Our study failed to show such a reduction. This may be explained by the fact that the proportion of patients undergoing radiography before the introduction of the OKR(phase1) was relatively low at 46.7%. When Steill et al. [4] showed a reduction in their study the baseline rate of radiography was 74%. Similar “post-rules” values were obtained with the previous study, showing a reduction to 53%,while in our study it went up to 50%. There was a significant reduction in the number of radiographs ordered on patients with negative OKR criteria, showing that the introduction of the rules did have an effect on the radiograph-requesting pattern of the doctors concerned. It also suggests that the doctors were comfortable with the use of the rules, and confident they could be used to exclude bony injury. Our study has several limitations which merit discussion. First, is the sample size. Because of the relatively small number of patients included and of diagnosed fractures, accuracy might have been overestimated. This study is a pilot study and can be used as a base to carry on another study to prove this fact. Multicenter studies included more 220 K. Nagpal, M. Marathe / Central European Journal of Medicine 2(2) 2007 216–221 patients are needed to conform our results. Another limitation of our study is that we did not perform follow-up of patients who did not receive radiography at the time of the ED visit. It is possible that not all patients with fractures were identified. We believe that this is unlikely. When telephone follow-up was performed in several similar studies, none of the patients in the no x-ray subgroups were identified as having fractures [2, 3]. We also did not test the intraobsever and interobserver reliability of these rules, which are relevant factors for broader use. The definition of any criterion might be differently understood by different physicians, whose skill for clinical examination might be very different. The ability to walk, for instance, might be difficult to assess in a reproducible way by two different physicians. Furthermore, this criterion is likely to be modified according to the time elapsed between trauma and clinical examination. In summary, the use of Ottawa rules did not allowed us to decrease the need for x-ray studies after isolated blunt knee trauma. More such studies are needed to confirm this.

References

[1] S.E. Gwilym, N. Aslam, W.J. Ribbbans and V. Holloway: “The impact of imple- menting the Ottawa ankle rules on ankle radiography request in A & E”, Int. J. Clin. Pract., Vol. 57, (2003), pp. 625–627. [2] I.G. Steill, G.A. Wells, I. McDowell, G.H. Greenburg, R.D. Mcknight, A.A. Cwinn, J.V. Quinn and A. Yeats: “Use of radiography in acute knee injuries: need for clinical decision rules”, Acad. Emerg. Med., Vol. 2, (1995), pp. 966–973. [3] I.G. Stiell, G.H. Greenberg, G.A. Wells, R.D. Mcknight, A.A. Cwinn, T. Cacciotti, I. McDowell and N.A. Smith: “Derivation of a decision rule for the use of radiography in acute knee injuries”, Ann. Emerg. Med., Vol. 26, (1995), pp. 405–413. [4] I.G. Stiell, G.H. Greenberg, G.A. Wells, I. McDowell, A.A. Cwinn, N.A. Smith, T.F. Cacciotti and M.L. Sivilotti: “Prospective validation of a decision rule for the use of radiography in acute knee injuries”, JAMA, Vol. 275, (1996), pp. 611–615. [5] I.G. Stiell, G.A. Wells, R.H. Hoag, M.L. Sivilotti, T.F. Cacciotti, P.R. Verbeek, K.T. Greenway, I. McDowell, A.A. Cwinn, G.H. Greenberg, G. Nichol and J.A. Michael: “Implementation of the Ottawa knee rule for the use of radiography in acute knee injuries”, JAMA, Vol. 278, (1997), pp. 2075–2079. [6] G. Nichol, I.G. Stiell, G.A. Wells, L.S. Juergensen, A. Laupacis: “An economic analysis of Ottawa knee rule”, Ann. Emerg. Med., Vol 34, (1999), pp. 438–447. [7] P.B. Richman, C.F. McCuskey, A. Nashed, S. Fuchs, R. Petrik, M. Imperato and X. Hollander: “Performance of two clinical decision rules for knee radiography”,J. Emerg. Med., Vol. 15, (1997), pp. 459–463. [8] D.C. Seaberg, D.M. Yealy, T. Lukens, T. Auble, S. Muthias: “Multicenter compari- son of two clinical decision rules for the use of radiography in acute, high risk knee injuries”, Ann. Emerg. Med., Vol. 32, (1998), pp. 8–13. K. Nagpal, M. Marathe / Central European Journal of Medicine 2(2) 2007 216–221 221

[9] S. Tigges, S. Pitts, S. Mukundan S Jr, Morrison D, Olson M, Shahriari A: “External validation of the Ottawa knee rules in ans urban trauma centre in United States”, Am. J. Roentgenol., Vol. 172, (1999), pp. 1069–1071.