VALIDITY OF OTTAWA KNEE RULES AT A TEACHING HOSPITAL OF EASTERN NEPAL

Bibhu Nath Mishra1*, Santosh Nepal1 , Surya Pd. Parajuli2 Received : 20 March, 2021 Accepted : 15 September, 2021 Published : 04 November 2021 ISSN: 2542-2758 (Print) 2542-2804 (Online) 1471 Birat Journal of Health Sciences Vol.6/No.2/Issue 15/May-Aug., 2021 Original Research Ar cle Mishra BN et al


Introduc on
Knee injuries are encountered frequently in Orthopedic emergency and Outpa ent departments. Radiographs are rou nely ordered in them, but not all of them demonstrate clear fractures. The decision for radiography based on subjec ve evalua on can help to reduce cost, decrease wai ng me, and unnecessary radia on exposure. We lack this informa on in our context.

Objec ve
The objec ve of this study was to find the validity of the O awa knee rule (OKR) in pa ents presen ng with acute knee injuries at a teaching hospital in eastern Nepal.

Methodology
A cross-sec onal study was conducted from March 2018 to February 2019 including 210 cases of acute knee injuries. The pa ents were evaluated as per OKR and their X-rays were evaluated too. Collected data were entered in MS Excel and analyzed by SPSS for validity.

Results
Out of the total of 210 eligible pa ents (122 males and 88 females) with a mean age of 43.97 years, the radiography rate was 100% but the yield rate was only 10.5%. Overall 69% of pa ents presented to the hospital within 24 hours of the injury and direct hit/trauma was the commonest mode of injury. Patella fractures were commonest followed by proximal bia fractures. There was a high sensi vity of 100% and a specificity of 42.02%. The rule yielded a Posi ve and Nega ve Predic ve value of 16.79% and 100%, respec vely. The OKR, if applied correctly, could result in radiography rate reduc on by 37.61%. The Fisher exact test result was significant at p<0.05.

Conclusion
OKRs is a valid tool to predict fractures in pa ents who has a history of acute knee injuries without chances of missing fractures. This rule can reduce unnecessary radiography in our setup as well.

INTRODUCTION
Acute knee injuries are very common injuries presen ng to [1][2][3][4] hospitals globally. Rou ne radiographs are ordered to rule out fractures in all but surprisingly fractures are reported in 2,5 less than 7% of cases only. It's well known that the majority of knee injuries have either meniscal or ligamentous ruptures but on contrary, the knee radiographs are done 3 more which demonstrate only fractures. Here comes the role of "clinical predic on rules" to improve diagnos c 6 efficiency. There are many clinical predic on rules developed at various ins tu ons by different groups for the determina on of the need for radiography to detect 1,8,9 fractures in pa ents presen ng with an acute knee injury. Most of them comprise of three or more variables obtained from the history, physical examina on, or simple diagnos c tests, and are developed a er original data collec on and 1,7-9 mul variate sta s cal analysis. The O awa Knee Rule (OKR) described by S ell et al. is a highly sensi ve and reliable clinical tool to determine the need for knee radiographs that have even proven to be 100% sensi ve for fracture predic on following knee injuries in 1,2 Canada. Few studies have reported that using these rules could reduce knee radiography by 28% without missing any 2,10 clinically significant fractures. This rule comprises five simple variables based on the history and physical examina on including age 55 years or older, isolated tenderness of the patella, tenderness at the head of the fibula, inability to flex the knee to 90 degrees, and inability to transfer weight for four steps both immediately a er the injury and in the 1,2,11 emergency department (figure 1).

Figure 1: O awa Knee Rule
According to OKR, acute knee injuries are categorized into two groups; those with at least one posi ve variable are considered likely to have a fracture and are advised for radiographs, whereas those without any posi ve variable are considered less likely to have a fracture and thus radiography can be safely deferred in them. This results in decreased radia on exposure, cost savings to some extent, and a reduc on in the number of nega ve knee radiographs 2,10-13 without adversely affec ng pa ent care as well.
.A meta-analysis of the results of six OKR valida on studies by Bachmann et al. showed a pooled sensi vity of 98.5% and 14 specificity of 49% for specific knee fractures. The authors, however, even recommended addi onal valida on studies.
In low socioeconomic countries like Nepal, this rule would be quite helpful to health personnel working at the periphery without X-ray facili es. It would help to decide whom to send for an X-ray thereby reducing unnecessary referrals. Even at urban health care centers with X-ray facili es, the decision for radiography based on subjec ve evalua on can help to reduce the cost, wai ng me, and unnecessary radia on exposure. Furthermore, the use of OKRs will help in the development of skill and a tude among doctors & other medical personals regarding the importance of clinical evalua on of pa ents and they will invest more me examining pa ents rather than wri ng radiological requisi on. However, in the context of eastern Nepal, we lack the valida on of this informa on. Hence, this study aims to find the validity of the O awa knee rule (OKR) in pa ents presen ng with acute knee injuries at a teaching hospital in eastern Nepal.

METHODOLOGY
A cross-sec onal study was conducted from March 2018 to February 2019 at Orthopedic department of Nobel Medical College Teaching Hospital located in eastern Nepal. A er taking ethical approval from the Ins tu onal Review Commi ee, 210 pa ents having acute knee injuries were enrolled consecu vely. Based on OKR, the pa ents with acute knee injuries were evaluated. The knee anatomy was broadly described to include the patella, the head and neck of the fibula, the proximal 8 cm of the bia, and the distal 8 cm of the femur. Blunt trauma was defined as any injury involving a direct trauma or mechanical force applied to the knee. Inability to bear weight was defined as the pa ent not being able to bear weight immediately (for four steps) a er the injury and in the emergency department (ED). The inclusion criteria included any pa ent of 18 years or above presen ng to either the outpa ent department (OPD) or the emergency department (ED) having acute knee pain within 7 days of injury and without any skin wound. Pa ents younger than 18 years, pregnant, referred from other hospitals with radiographs, more than 7 days old knee injuries, mul ple injuries, and paraplegics were excluded. A ending physicians were trained to assess the pa ents as per the OKR. The finding was recorded in specifically designed pro forma and ordered for knee radiographs as usual prac ce, independent from OKR. All the pa ents had at least two views taken in their radiographs; anteroposterior and lateral, whereas the tangen al (skyline) view of the patella was at the discre on of the trea ng physician. The radiographs were reviewed by the trea ng physician for immediate treatment and were interpreted, within 24 hours, by an orthopedic surgeon who had been blinded to the contents of the data collec on sheet. The results of the radiographic examina on were recorded on the data collec on sheet. Now, the pa ents were divided into two groups; OKR posi ve and OKR nega ve groups, and their X-ray findings as fracture were further divided into two sub-groups; X-ray posi ve and X-ray nega ve groups. The univariate analysis was performed for frequency and percentage and bivariate analysis was performed for fisher exact test and validity.

RESULTS
Among 210 pa ents, there were more males (58%) than females (42%). The age ranged from 20 years to 74 years with a mean age of 43.97 years. Regarding mode of injury, direct trauma (fall/object impact) was the commonest cause (48.1%), followed by a twis ng knee injury (table 1). More cases presented to ED (53%) in comparison to OPD (47%) and 69% of the pa ents presented within 24 hours of sustaining the injury. Both sides were almost equally affected with a slight predominance of the right side (52%). There were fewer numbers of fractures overall (10.5%) of which patella fractures (5%) were the most common followed by proximal bia fractures (table 2 and figures 2a-2e).   Since the sample size was rela vely smaller, so we used the Fisher exact test with a 2 x 2 con ngency table (table 3). The Fisher exact test result was significant at p<0.05.   (Figure 3). As shown in table 3, the sensi vity was 100% and the specificity was 42.02%. The Posi ve Predic ve Value (PPV) was 16.79% and the Nega ve Predic ve Value (NPV) was 100%, with an X-ray reduc on rate of 37.61%. The100% sensi vity ensures that none of the pa ents with fractures were missed and the Nega ve Predic ve value of 100% signifies that the chance of observing clinically significant fractures in pa ents who were diagnosed as OKR nega ve was zero. The high X-ray reduc on rate signifies that unnecessary radiography was done in a lot of pa ents. But the pa ents and their radiographs were evaluated by different persons, so chances of inter-observer and intra-observer varia ons exists and we accept it as one of the limita ons of our study.

DISCUSSION
Ian G S ell and colleagues performed the first study among 1054 adult pa ents in 1995 at two university hospitals' emergency departments in O awa, Canada to derive a highly sensi ve clinical rule for fracture predic on following 1 acute knee injuries. This decision rule had a sensi vity of 1.0, a specificity of 0.54 for knee fractures iden fica on, and accounted for rela ve reduc on by 28% for radiography use as well. They had localized bony tenderness over the patella or the head of the fibula or were unable to bear weight both immediately a er the injury and in the emergency department. These features were called the O awa knee Rules (OKRs). Soon a er that, S ell et al. performed a second study to validate and refine the 2 previously derived OKR. The study again showed that the decision rules had a sensi vity of 100% with a 28% rela ve reduc on in the use of radiography. Therea er, Steill et al. conducted an interven onal study to assess the impact on 5 the clinical prac ce of implemen ng the OKR. They found that there was a rela ve reduc on of 26.4% in the propor on of pa ents referred for knee radiography in the interven on group (77.6% vs 57.1%; P < 0.001), but a rela ve reduc on of only 1.3% in the control group (76.9% vs 75.9%; P = 0.60). These changes over me were significant when the interven on and control groups were compared (P < 0.001). The rule was found to have a sensi vity of 1.0 (95% CI, 0. Similarly, the specificity matching in accordance (42.02%) to our study has been 2,5,12,19 reported by few other studies as well. According to our study, implemen ng the OKR would have resulted in an overall reduc on in the use of radiographs by 37.61%. However, the true reduc on in radiography rates cannot be determined unless an interven onal trial is performed. The poten al reduc on in the number of requested knee radiographs that could be achieved in the emergency department or the outpa ent department (OPD) by applying the OKR is slightly lower than the reduc on es mated in a study performed by Emparanza

CONCLUSION
The commonest mechanism of knee injury was direct trauma and it occurred mostly in young adults. O awa Knee Rules were found to be accurate and highly sensi ve tools for the predic on of fractures in pa ents with acute knee injuries. So, with its correct implementa on, the chances of missing fractures are almost zero and have the benefits of reducing treatment cost and radia on exposure.

LIMITATIONS OF THE STUDY
Our study has few limita ons. Inter-observer biases cannot be ruled out. Even the chances of selec on bias exist as enrolment in this study was based on the willingness of pa ents, their rela ves, and the co-opera on of the duty doctors too.

FINANCIAL DISCLOSURE
None