Open Access Open Badges Research article

The role of patient expectations in predicting outcome after total knee arthroplasty

Anne F Mannion12*, Stephane Kämpfen3, Urs Munzinger3 and Ines Kramers-de Quervain4

Author Affiliations

1 Department of Research and Development, Schulthess Klinik, Lengghalde 2, 8008 Zürich, Switzerland

2 Department of Rheumatology and Institute of Physical Medicine, University Hospital Zürich, Gloriastrasse 25, 8091 Zürich, Switzerland

3 Department of Lower Extremity Orthopaedic Surgery, Schulthess Klinik, Lengghalde 2, 8008 Zürich, Switzerland

4 Department of Rheumatology, Schulthess Klinik, Lengghalde 2, 8008 Zürich, Switzerland

For all author emails, please log on.

Arthritis Research & Therapy 2009, 11:R139  doi:10.1186/ar2811

Published: 21 September 2009



Patient's expectations are variably reported to influence self-rated outcome and satisfaction after medical treatment; this prospective study examined which of the following was the most important unique determinant of global outcome/satisfaction after total knee arthroplasty (TKA): baseline expectations; fulfilment of expectations; or current symptoms and function.


One hundred and twelve patients with osteoarthritis of the knee (age, 67 ± 9 years) completed a questionnaire about their expectations regarding months until full recovery, pain, and limitations in everyday activities after TKA surgery. Two years postoperatively, they were asked what the reality was for each of these domains, and rated the global outcome and satisfaction with surgery. Multivariable regression analyses using forward conditional selection of variables (and controlling for age, gender, other joint problems) identified the most significant determinants of outcome.


Patients significantly underestimated the time for full recovery (expected 4.7 ± 2.8 months, recalled actual time, 6.1 ± 3.7 months; P = 0.005). They were also overly optimistic about the likelihood of being pain-free (85% expected it, 43% were; P < 0.05) and of not being limited in usual activities (52% expected it, 20% were; P < 0.05). Global outcomes were 46.2% excellent, 41.3% good, 10.6% fair and 1.9% poor. In multivariable regression, expectations did not make a significant unique contribution to explaining the variance in outcome/satisfaction; together with other joint problems, knee pain and function at 2 years postoperation predicted global outcome, and knee pain at 2 years predicted satisfaction.


In this group, preoperative expectations of TKA surgery were overly optimistic. The routine analysis of patient-orientated outcomes in practice should assist the surgeon to convey more realistic expectations to the patient during the preoperative consultation. In multivariable regression, expectations did not predict global outcome/satisfaction; the most important determinants were other joint problems and the patient's pain and functional status 2 years postoperatively.