Are the Pittsburgh Knee Rules reliable at identifying patients who require radiography?

Date First Published:
July 8, 2011
Last Updated:
July 11, 2011
Report by:
Bryony Patrick, Medical student (University of Manchester)
Search checked by:
Bryony Patrick, University of Manchester
Three-Part Question:
In [adult patients with traumatic knee pain] are the [Pittsburgh Knee Rules] reliable at [identifying patients who require radiography]?
Clinical Scenario:
A 30 year old man presents to the Emergency Department with a painful, swollen knee after sustaining a sports-related injury. You suspect a soft-tissue injury and wonder if he still requires an x-ray?
Search Strategy:
Medline 1948 – July 2011 using the Pubmed interface
Embase 1980 – Week 26 2011
Search Details:
(clinical[All Fields] AND decision[All Fields] AND rules[All Fields] AND ("knee"[MeSH Terms] OR "knee"[All Fields] OR "knee joint"[MeSH Terms] OR ("knee"[All Fields] AND "joint"[All Fields]) OR "knee joint"[All Fields])) AND ("humans"[MeSH Terms] AND English[lang] AND "adult"[MeSH Terms]).
pittsburgh[All Fields] AND ("knee"[MeSH Terms] OR "knee"[All Fields] OR "knee joint"[MeSH Terms] OR ("knee"[All Fields] AND "joint"[All Fields]) OR "knee joint"[All Fields]) AND rule[All Fields]
‘Pittsburgh knee rule’ limited to related terms
Outcome:
Found 8 papers that were relevant of which 4 were suitable for appraisal.
Relevant Paper(s):
Study Title Patient Group Study type (level of evidence) Outcomes Key results Study Weaknesses
Clinical decision rule for knee radiographs Seaberg D.C., Jackson R. 1994 USA 133 patients with acute knee injuries assessed for positive signs outlined in clinical decision rule to predict need for x-ray (fall / blunt trauma mechanism AND age <12 or >50 years
OR inability to ambulate) designed from initial retrograde study of 201 patient records. All underwent radiography to assess sensitivity and specificity of rule.		 Decision rule / cohort Small sample sizes. Not all potential predictors included during phase 1. Focus on patients who needed x-rays rather than patients that didn’t. 2/3 of patients with fractures had multisystem trauma.
A clinical decision rule in the evaluation of acute knee injuries Bauer S.J., Hollander J.E. et al 1995 USA 207 patients with acute knee injuries assessed with standardised history and examination tool. Physician predicted likelihood of fracture before radiography. Most underwent radiology to determine sensitivity and specificity of rule and likelihood of fracture from examination. Decision rule / cohort Sensitivity and specificity of criteria from history and examination associated with higher incidence of fracture. 100% sensitivity and 100% negative predictive value for fracture with inability to weight bear OR effusion OR ecchymosis OR severe joint line tenderness OR severe localised swelling. Small sample size. Potential selection bias as not all patients x-rayed. Clinician’s compliance with standardised tool not assessed.
Sensitivity and specificity of clinicial prediction of fracture. 56% sensitivity and 91% specificity of physician prediction of fracture
The Pittsburgh decision rule: triage nurse versus physician utilization in the Emergency Department Simon L.V., Matteucci M.J. et al 2006 USA 152 consecutive patients with acute knee injuries assessed by triage nurse and physician using Pittsburgh rule. All underwent radiology. Level of inter-observer agreement calculated. Cohort Kappa coefficient for need for x-ray K = 0.83 ‘almost perfect’ for overall need for x-ray Small sample size. Patient group included children. Higher fracture rate than other studies (8.6%)
Kappa coefficient for blunt trauma/fall K = 0.67 ‘substantial’
Kappa coefficient for age <12 or >50 years K = 1.0 ‘almost perfect’
Kappa coefficient for ability to walk 4 weight-bearing steps in ED K = 0.67 ‘substantial’
Sensitivity and specificity of rule Sensitivity 77%, specificity 57% (physicians), 58% (nurses)
Clinical decision rules discriminate between fractures and nonfractures in acute isolated knee trauma Weber J.E., Jackson R.E., et al 1995 USA 242 patients with acute knee injuries (less than 24 hours old) assessed for lengthy standardised history and examination criteria. Physician predicted likelihood of fracture before x-ray. Most underwent radiography to determine sensitivity and specificity of criteria, and clinician’s prediction of fracture. Decision rule Sensitivity and specificity of criteria individually and combined. Age over 50 crude odds ratio = 4.0 Small sample size. Potential selection bias as not all underwent radiography. Patients excluded if knee injury sustained >24hours previously. Subjective criteria used, eg severity of effusion. Clinician’s compliance with standardised tool not assessed.
Sensitivity and specificity of rule criteria (ability to walk without limping AND twist injury without effusion Sensitivity = 0.86, specificity = 0.83, misclassification rate = 58%
Author Commentary:
Limited number of papers on this topic area, all with very similar study designs. A meta-analysis study is required to target small sample sizes and show which criteria have the highest specificity and sensitivity.
Bottom Line:
The Pittsburgh knee rule is a sensitive and specific decision rule for identifying patients who have signs and symptoms associated with fracture. It is a useful tool for use in the Emergency Department to help clinicians decide if a patient requires an x-ray.
References:
  1. Seaberg D.C., Jackson R.. Clinical decision rule for knee radiographs
  2. Bauer S.J., Hollander J.E. et al. A clinical decision rule in the evaluation of acute knee injuries
  3. Simon L.V., Matteucci M.J. et al. The Pittsburgh decision rule: triage nurse versus physician utilization in the Emergency Department
  4. Weber J.E., Jackson R.E., et al. Clinical decision rules discriminate between fractures and nonfractures in acute isolated knee trauma