Introduction

Physical activity has been proven to prevent several chronic diseases, improve fitness, and decrease mortality [1, 2]. Regular exercise producing cardiovascular fitness, more over has been shown to stimulate weight loss, reduce anxiety as well as depression, and improve bone density and muscle coordination [3–6].

For older adults, staying physically active is of utmost importance, which is also true for the increasing number of patients having total knee or hip replacement.

Because of the beneficial effects of physical activity, international guidelines have been developed for levels of health-enhancing physical activity. These guidelines recommend 30 min or more of moderate-intensity aerobic (endurance) physical activity at least 5 days/week or vigorous-intensity aerobic physical activity for a minimum of 20 min at least 3 days/week [7, 8]. However, the current question of most interest always will be: How much loading will an artificial knee or hip tolerate? What kind of physical activity should be recommended to those patients? As the design and quality of the prostheses are currently improved, the optimal answer will have to be altered, which was illustrated by surveys among American hip and knee surgeons in 2001, 2007 and 2008 [9–11]. From these surveys, the then current specialist’s recommendations were derived. Aside from those; no recommendations have been published to our knowledge, and national or international guidelines have not been developed. Hence, updated specialist recommendations are currently warranted. In the present study, we have investigated the current Danish specialist’s recommendations.

Methods

From the Danish National Hip Register (DHR) and the Danish National Knee Register (DKR) 2010 reports, we selected all departments, both public and private, performing at least 100 THAs or TKAs. The leading surgeon in each department was asked to fill in a questionnaire about which athletic activities patients in his/her department were allowed to take part in after uneventful recovery from surgery. In Denmark, most arthroplasties are performed in specialised elective units manned with 2–6 surgeons who perform between 80 and 300 arthroplasties each per year. Furthermore, all Danish hospitals (private and public) have been accredited according to The Danish Healthcare Quality Programme [12], a public accreditation programme working by the same principles as The Joint Commission and ISQua. Accreditation is achieved only if the control panel detects full consistency between treatment descriptions on the hospital’s website, in the written instructions given to the patients, and in daily practice at the clinic. Regardless of whether the latter is provided by the department head or the youngest specialist.

The selection of activities in the questionnaire is based on the 2005 survey from the Danish Institute of Governmental Research [13], from which we selected the 31 most popular sports activities in the age-group 60–69. For each activity, the surgeon should decide whether the fully recovered patient was advised to: 1—take part regardless of previous experience with the activity; 2—take part if the patient had experience with the activity before surgery; or 3—not take part in the activity. If the questionnaire was not returned within 2 months, we sent a reminder email, and if necessary a complimentary phone call was used to heighten the response rate.

Statistical method

We compared the percentage of surgeons allowing each activity based on our questionnaires to those of Klein (2007). Similarly, we compared the ‘allowed with experience’ and ‘not allowed’ categories.

In the following, assume that a certain allowance category (‘allowed’, ‘allowed with experience’, or ‘not allowed’) and an impact (‘high’, ‘intermediate’, and ‘low’) or a combination of these impacts (‘all’ and ‘high or intermediate’) are considered for statistical analysis.

Let m be the number of responding surgeons in Klein (2007) and n the number of responding surgeons in DK (2012). For activity number j (for example alphabetic order), let x j/m be the percentage of Klein (2007) surgeons recommending ‘Yes’ and y j/n the percentage of DK (2012) surgeons recommending ‘Yes’. As we assume that the surgeons are independent and also that the activity recommendations within a surgeon are independent, we can analyse the percentage using a logistic regression model with activity and source [Klein (2007) or DK (2012)] as explanatory variables. If we only had one activity, this would correspond to comparing two binomial proportions.

If the source variable is statistical significant (p < 0.05), this can be interpreted as an overall difference between Klein (2007) and DK (2012). This difference between sources can be converted into an odds ratio that can be interpreted as the increase in recommendation percentage (averaged over all activities) if surgeons from DK (2012) were to make a recommendation instead of surgeons from Klein (2007).

We only included the activities where both sources [Klein (2007) or DK (2012)] had a recommendation. Later, when comparing THA with TKA as sources, the same type of analysis was conducted.

Results

Twenty seven departments for TKA and 28 for THA met the inclusion criteria, and 20 respective 25 filled in the questionnaire. This gives us a response rate of 74 % for TKA and 89 % for THA. Calculated from the number of procedures performed at each department, the response rates are 78 % for TKA and 91 % for THA.

Over the years 1999–2007, there was a tendency for the US recommendations to be more liberal,and for instance, single tennis was not allowed before 2007 for THA (Table 1) and bicycling and canoeing was not allowed after TKA before 2005 (Table 2). Compared with the US recommendation in 2007, the Danish recommendations in 2012 are significantly more liberal. A total of 87 % allowed any athletic activity in 2012 compared to 77 % in 2007 (p < 0.0001, Table 3). Even high-impact activities were allowed by 48 % in 2012, compared to 21 % in 2007. Of these 48 %, 13 % allowed high-impact activities only to experienced patients. In 2012, still 13 % of surgeons did not allow athletic activity to THA patients, compared to 21 % in 2007.

Table 1 Expert opinions for THA
Full size table
Table 2 Expert opinions for TKA
Full size table
Table 3 Statistics: results for comparison of Klein (2007) with DK (2012) (Table 1) and THA with TKA (Tables 1, 2)
Full size table

Compared with THA (Table 1), the recommendations for TKA (Table 2) were significantly more restrictive (Table 3). Twenty per cent did not allow any athletic activity after TKA compared to 14 % after THA (p < 0.0001). Whereas 55 % allowed high-impact activities after THA only 38 % allowed this after TKA. Among these, only 22 % allowed high-impact activities if not experienced, compared to 35 % for THA. The recommendation by the departments did not depend on the number of procedures performed (Fig. 1). There was a significant difference between Klein (2007) and DK (2012).

Fig. 1
figure 1

Danish experts recommendations plotted against their production

Full size image

Discussion

Our results could be biased by the chosen method; data collected are in fact the view of the heads of department in the larger arthroplasty units in Denmark, but we believe that all answers are given in the spirit of each departments fixed (and accredited) protocol for sporting activities, for all the activities covered and adhered to by all relevant specialists in the department. It is not possible for a senior member of a Danish orthopaedic department to maintain a very specialist practice run against differing views to his/her colleagues. In fact, all arthroplasty surgeons within a unit are forced to consensus on a fixed protocol for all treatment aspects. Otherwise, accreditation will be redrawn.

The orthopaedic literature on athletic activity after THA and TKA is mostly limited to small retrospective studies with short-term follow-up [14, 15]. One longitudinal four-centre study in 2005 found that most patients with hip (97 %) and knee (94 %) OA had performed sports activities during their life, only 36 % (hip patients) and 42 % (knee patients) had maintained sports activities at the time of surgery. Five years post-operatively, the proportion of patients performing sports activities increased to 52 % among patients with THA, but further declined to 34 % among those with TKA. Accordingly, the proportion of patients with THA performing sports activities for more than 2 h a week increased from 8 to 14 %, whereas this proportion decreased from 12 to 5 % among patients with TKA [16]. Differences in pain 5 years after joint replacement might explain some of the differences of sports activities as persistent post-operative pain were reported by 9 % of patients with THA and by >16 % with TKA. Also, the surgeons’ belief with regards to the mechanical strength of the artificial joint and its surroundings and the risks of overuse and traumatic injury, and consequently, recommendations are probably highly decisive for the patient. The present study is the first to demonstrate that the recommendations for THA patients are significantly more liberal compared with TKA, a total of 20 % of the departments did not allow any athletic activity after TKA, and this circumspection is probably only based on personal experience. To our knowledge, evidence to conclude increased soreness of TKA compared with THA has not been published. When patients who have undergone joint replacements choose to participate in athletic activity, orthopaedic surgeons should provide information with which to evaluate the risk of sports activity and recommend appropriate athletic activity. A trend among experts to allow more athletic activity and relax restrictions of sports activity after joint replacements may be based on outstanding patient outcomes, increasing surgeon confidence in surgical technique, and innovations in joint implants. This trend may also be a response to patients’ demands to participate in athletics after surgery. However, this trend in expert opinion is not evidence-based and may not be in the best interest of patients.

At this state, we still linger at the lowest possible level of evidence: “the expert opinion”. An increasing number of experts allow for increasing activity—but that will never alter the evidence level. Future research combining large patient cohorts with yearly Patient Reported Outcome Measures (PROM) regarding activities and level will give us the possibility to tell whether there are any differences in implant survival between patient groups in different sports and activity levels and possibly tell whether special implants should be preferred for patients with special demands for their future sports activities.

Discussion of statistical method

As several similar analyses have been conducted in Table 3, and some analyses of these have a subset of data in common, the reader should be cautious to interpret them jointly rather than as one single analysis at a time. Some multiple-testing correction could be invoked, but as it is not obvious how, as only a subset of data is in common, we chose not to perform such a correction.

It would be relevant to have the all the original responds from the surgeons in Klein (2007) in order to analyse if there were within-surgeon structure. We could have done this type of analysis for comparing THA with TKA as we have all the original observations, but to maintain the same interpretation of the results between the two analyses, we chose not to.