56 / 225
56
· DOS Abstracts
A new screening algorithm to improve the referral pattern of
outpatient orthopedic knee patients. Development and evalua-
tion based on patient-reported data and radiographs.
Lone Ramer Mikkelsen, Mette Garval, Carsten Holm, Søren Thorgaard Skou
Interdisciplinary Research Unit, Elective Surgery Centre, Silkeborg Regional Hospital;
2. Department of Physiotherapy, Elective Surgery Centre, Silkeborg Regional Hos-
pital; Elective Surgery Centre, Silkeborg Regional Hospital; Department of Physio-
therapy and Occupational Therapy AND Research Unit for Musculoskeletal Function
and Physiotherapy, Institute of Sports Science and Clinical Biomechanics, Næstved-
Slagelse-Ringsted Hospitals AND University of Southern Denmark
Background:
Many knee patients referred to outpatient orthopedic clinics (OOC)
are not (yet) candidates for surgery and might benefit from conservative treatment.
If it is possible to identify relevant patients to refer to the orthopedic surgeon (OS)
it could potentially improve efficiency and quality of care in the OOC.
Purpose / Aim of Study:
To develop and test a screening algorithm to define ap-
propriateness of referral to OS based on pre-visit patient-reported outcomes and
radiographical findings thereby being applicable prior to clinical examination.
Materials and Methods:
Prior to clinical examination, 173 consecutive patients
with a first-time referral to the OOC completed questionnaires, and radiographic
osteoarthritis severity was graded. The gold standard for relevant referral to the
OS was based on actual treatments, referral to other medical specialists or further
diagnostics. The performance of the algorithm in predicting relevant referrals and
total knee replacement (TKR) was assessed using sensitivity, specificity, positive
predictive value (PPV) and negative predictive value (NPV).
Findings / Results:
Of the 173 patients, 40% (n=69) underwent TKR and further
25% (n=44) were considered relevant to refer to OS due to other reasons than
surgery. Sensitivity, specificity, PPV and NPV for prediction of relevant referral to
OS were 0.70, 0.56, 0.76 and 0.48, respectively. The corresponding performance
estimates for prediction of TKR surgery were 0.92, 0.56, 0.55 and 0.92.
Conclusions:
The algorithm was able to identify most patients relevant to refer
to OS, but was less suitable for identifying those not relevant. The algorithm per-
formed excellent in predicting TKR surgery. With further development, this screen-
ing algorithm might be able to improve the referral pattern and thereby improve
patient care and efficiency in the OOC.
No conflicts of interest reported
8.