RADIOGRAPHIC FOLLOW-UP FOR ASYMPTOMATIC PATIENTS AFTER PRIMARY CEMENTED TOTAL KNEE ARTHROPLASTY

Participants: R.B. Makim, T. O'Keefe, L. S. Matthews, M. L. Greenfield

Keywords: total knee arthroplasty, radiographs, alignment, asymptomatic osteolysis

Introduction

Total knee arthroplasty (TKA) is one of the most commonly performed operations in orthopaedic surgery. Approximately 150,000 total knee arthroplasties are performed annually in the United States. The use of plain film radiography to evaluate the arthroplasty in follow up is very common. Unfortunately, there have been few reports in the literature regarding guidelines for radiographic follow-up. Many surgeons obtain x-rays at defined intervals, such as immediately postoperatively, at 6 weeks, at 3 months, at 6 months, and thereafter, annually or bi-annually. Significant time and cost savings may be accomplished if these x-rays can be shown to be unnecessary.

X-rays are valuable in determining alignment of the components and of the femur and tibia. X-rays can also demonstrate loosening of the components, which can be seen as abnormal, progressive radiolucent lines, or a shift in the components. With early polyethylene-induced aseptic loosening, the patient may be asymptomatic, but usually reports symptoms of effusion, stiffness, and/or pain that results in a clinic visit to the physician. Generally, those patients reporting these symptoms are bothered persistently and may be functionally impaired to varying degrees. There is no question that radiographs are necessary in this situation.

X-ray abnormalities can occur in asymptomatic patients following total joint arthroplasty. In total hip arthroplasty, patients can be asymptomatic with moderate or even large osteolytic lesions. Cemented TKA's differ from total hip arthroplasties in that osteolysis following the former procedure is an extremely rare phenomenon. Detecting osteolysis early in the total hip is of benefit since revision arthroplasty is more easily and better accomplished before bone loss becomes severe. In the cemented total knee, however, its incidence is so low that routine attempts at detection could be considered cost-prohibitive.

Other x-ray abnormalities, including radiolucent lines as discussed previously, femoro-tibial alignment, component-shaft alignment, osteonecrosis, and fractures are significant entities but do not alter treatment if the patient is asymptomatic. If the patient is not complaining of persistent pain, instability, stiffness, or effusion that limits their desired lifestyle, revision surgery is not indicated. Therefore, in the asymptomatic patient following cemented primary total knee arthroplasty, x-rays have questionable benefit. The purpose of this study is to examine this issue and put forth some guidelines regarding radiographic follow up for these patients.

The overall goal of this study is to ascertain whether follow-up x-rays of asymptomatic patients following primary cemented total knee arthroplasty are of value. Specifically, we will determine the number of asymptomatic patients with radiographic abnormalities that may have influenced or changed care over a 3-5 year period.

Materials and Methods

Approximately 200-250 consecutive patients following TKA will be selected between the years of 1989 and 1992. Both males and females of all ages, races, and diagnoses will be included.

Inclusion criteria: All patients of Dr. Okeefe's practice who received a primary cemented TKA between the years of 1989 and 1992.

Exclusion criteria: Inadequate radiographic follow-up: X-rays must have been taken postoperatively (within 3 months of the operation) and subsequently at least 3 years postoperatively.

Symptomatic patients (with early persistent symptoms of pain, effusion, instability, or activity-limiting stiffness 3 months or later after the operation.)

Patients must have had at least 3 post-operative visits after primary TKA.

A retrospective study will be performed with univariate statistics to summarize means and frequencies. Data from the charts will include age, diagnosis, date of operation, side of TKA, date of revision surgery, and presence or absence of symptoms (persistent pain, instability, effusion, or activity-limiting stiffness) at 3 months, 6 months, 1 year, and yearly thereafter for at least 2 additional years. Data from radiographs will include femorotibial alignment (varus or valgus), component-shaft alignment, presence or absence of radiolucent lines, and if present, progression of the lines. The measurements will all be made using the same goniometer and ruler. A kappa statistic will be obtained to verify interrater and intraobserver reliability. (This reliability study analysis will consist of information obtained from approximately 25 patients. The same process will be repeated by one of the secondary authors, and then again by the primary investigator a week or so later.)

The major outcome variable will be a change in patient treatment as noted in the patient record. We will determine if any treatment changes were made on the basis of abnormal x-rays in asymptomatic patients. We can then conclude whether there was any benefit in obtaining these x-rays. Potential confounders include: inadequate documentation in the chart, inconsistencies in the taking of the x-rays (angle of beam, magnification, degree of knee flexion/extension), and differences in the surgery itself (type of implant used, technique of implantation and cementation).

In addition to reporting simple univariate statistics, survival analysis based on a 3-5 year survival of the implants in asymptomatic patients will be conducted.

Progress

The data are currently being collected.