Difference in early all-cause mortality among patients having hip arthroplasty a Swedish perioperative registry study 2013–2022

As anticipated, a significant disparity in mortality rates was observed between the two primary cohorts: those with osteoarthritis and those with fractures. This difference was largely attributed to variations in demographics, specifically age and ASA-class. The mean age difference of twelve-years and a 40% discrepancy in the proportion of patients classified as ASA 3–5 is most certainly of major importance. Unfortunately, the SPOR register does not encompass fragility scores, which could provide further insights into these differences,

The hip fracture group was associated with a 21.5 higher odds ratio for all cause 30-day mortality as compared with the osteoarthritis group in the adjusted model. The mortality among the fracture patients is similar to what was found in an earlier study, assessing all hip fracture surgery 2016 and 2017 [25]. The mortality following hip arthroplasty in the present study is also in line with a study by Maceroli et al. from US [26]. They found a 30-day mortality after hemiarthroplasty of 8.4% compared with 5.7% after total hip arthroplasty in hip fracture patients above 60 years of age. The present results in terms of the hip fracture mortality are also similar to the results in a Norwegian study that assessed the impact of a fast-tracking concept in hip fracture patients [27]. These researchers found a 7.9% versus 6.5% 30-day mortality for the conventional and fast-tracked groups respectively. It should be acknowledged that only approximately 50% of the patients underwent an arthroplasty procedure, either hemi- or total arthroplasty.

A study by Pedersen at al. based on the Nordic arthroplasty register assessing mortality following total hip arthroplasty for primary osteoarthritis found a 30-day mortality in the 188,606 patients included to be 0.21% for cemented total hip arthroplasty and 0.12%) for cementless fixation. This is similar to the osteoarthritis cohort in the present study (0.12%) [12], . Large differences in the one-year mortality between patients having undergone arthroplasty for either osteoarthritis or hip fracture were also observed in a recent German Arthroplasty-based study by Szymsky et al. [17], 12.6% compared with 3% respectively.

There are several potential factors contributing to the large difference in early mortality. Age and ASA-class are of course of importance, but fragility is most certainly an important factor in combination with the trauma as such and co-morbidity. It is also important to recognize that the mortality rates associated with osteoarthritis may be somewhat overstated. This potential discrepancy could stem from the fact that some private units, which likely treat a healthier cohort of patients, may not fully report their data to the SPOR register, particularly in the earlier time period of this cohort. Consequently, the observed difference in mortality rates between patients with fracture indications and those with osteoarthritis could, in reality, could be even more pronounced.

We found that a majority of arthroplasties performed in fracture patients were hemi arthroplasty. The hemi arthroplasty was among these fracture patients associated with a higher all-cause mortality, and the odds ratio for all-cause 30-day mortality was still significantly higher as compared to total and hybrid arthroplasty after adjusting for age and ASA-class, anaesthetic and fixation technique (OD 2.2). This is a finding in line with the study by done on the Swedish hip register published in 2018 [28].

The study by Pedersen et al. [12] study found no significant differences in mortality between cemented and cementless arthroplasty in patients with osteoarthritis. This is in harmony with the present study. A study by Ekman et al. [10] from Finland including 62,221 hip arthroplasties for osteoarthritis found in a similar manner no differences in early and postoperative mortality between cemented and cementless fixations. Lindberg-Larsen et al. [18] studied patients with osteoarthritis above 70 years of age based on data from the Danish hip arthroplasty register. They found similar risks of mortality within 30 days (0.2% vs. 0.3%) in these elderly patients. Ricksten et al. [14] studied patients having undergone hip arthroplasty due to femoral neck fractures and found a significantly different 30-day mortality, with 3% in the cementless group and 9% in the cemented group (p = 0.03). This is also in harmony with the results of the present study. Tsai et al. [19] studied patients aged 80 and above who underwent cemented hip arthroplasty and they reported a higher 30-day mortality. Interestingly, Kristensen et al. [13] in a Norwegian register-based study found no mortality differences at one-year between cemented and cementless arthroplasties after hip fractures. Fernandez et al. [15] presented results from a prospective randomized trial that compared cemented and cementless hemiarthroplasties in 1,215 patients 60 years and above with hip fracture. They found modest, but significantly better quality of life and lower risk for periprosthetic fractures in the cemented group. The one-year mortality was 23.9% in the cemented group and 27.8% in the uncemented group. Springer et al. [16] and Kheir et al. [29] found likewise in register-based studies from US a clear reduction in the risk of revision in patients having undergone cemented arthroplasty for hip fracture patients. Springer et al. also found that stem cemented fixation was associated with an increased odds ratio of in-hospital, 90-day and one-year mortality.

The most similar study to the present one was published by Dale et al. [11] covering the time period 2005 to 2018, using data from the Norwegian Arthroplasty Register. It included in a similar manner both osteoarthritis and hip fracture patients, and different fixation techniques. They found that all modes of fixation had similar adjusted 3-day, 30-day, 90-day mortality risks. The only difference compared with the present study were the lower odds ratio for hybrid fixation, as shown by the regression analysis. These results should, however, be put into perspective, considering that the hybrid fixation technique in the present study was used in only a limited proportion of patients, i.e., only 9% in the entire cohort.

There were no differences in mortality associated with anaesthetic techniques after adjustment for age, ASA, indication and fixation technique. This finding is in line with a previous study assessing mortality after hip fracture surgery based on SPOR data [25]. The International Consensus on Anaesthesia-Related Outcomes after Surgery group (ICAROS) published a systematic review and meta-analysis in 2019 [30]. The consensus statement, based on the meta-analysis, was a recommendation for neuraxial anaesthesia for hip arthroplasty, but did also comment that the evidence level was moderate to low. The REGAIN study published 2021 including patients 50 years of age or older who underwent surgery for hip fractures at 46 U.S. and Canadian hospitals, did not find any differences between general and spinal anaesthesia, with respect to survival and recovery of ambulation at 60 days [31]. A more recent systematic review studied the anaesthesia impact on multiple outcomes after hip fracture surgery. This study did now show any differences between spinal and general anaesthesia, except for patients with renal impartment. Mortality did not differ either [21].

The present study is based on the Swedish perioperative register SPOR. It covers the period 2013 until the first half year of 2022. It includes all primary hip arthroplasties operated on due to osteoarthritis or hip fractures, age 18 years or more during the period covered. It covers patients operated on in university hospitals as well as regional and standard hospitals, thus both in high volume centres and low volume centres. The mortality is based on the collective consignment between the SPOR and Swedish death register and should thus present robust mortality data. One should also keep in mind that Sweden has a tax funded health care system and patients of all ages, and with comorbidities and sociodemographic background undergo surgery on indication-based premises.

There are also limitations to this study. There was a surprisingly high number of missing data related to the anaesthetic technique. We did not include the time to surgery for the hip fracture group of patients or to what extent enhanced recovery after surgery protocols have been implemented.

The present study shows a large difference in early mortality between the two main subgroups, osteoarthritis, and fracture patients. We had an interest in assessing current early mortality and how much of a difference that could be seen between the two major groups of hip arthroplasty in Sweden and the impact of cement use. The early mortality was expectedly associated with higher age and ASA-class, while anaesthetic technique and fixation appear to have minor impact. The more than 8% 30-day mortality among the hip fracture patients, although in line with mortality seen among hip fracture patients, calls for a deeper analysis. It should also be acknowledged that the fracture group of patients include 50 patients with ICD-10 codes (723, 7230, 7231, 724, 7270, 728, 7280, 729, and 7290) covering any types of femur fractures, but all of these patients had a hip arthroplasty. Whether these patients had undergone also more distal procedures cannot be assessed as further procedural codes were not available in the data set.

The register does not contain information about medical history and current medications and information about the cause of death is not available. The register does not contain information related to complications during the hospital stay after discharge from the recovery room and information about return to hospital needs. Assessment of fragility and activities of daily living is also lacking. The combination of advanced age, a high ASA-class, and fragility undeniably plays a critical role in determining mortality outcomes following intermediate surgical procedures like hip arthroplasty. This triad significantly impacts patient prognosis and highlights the need for meticulous preoperative assessment and optimisation. Therefore, it is of great importance to further analyse patient factors and if possible, improve the perioperative care. The choice of hemi vs. total arthroplasty for the hip fracture patients’ needs also further studies. The increased early mortality seen in the present study must be put in context of the benefits associated with in use in a longer context [32].