Internal fixation versus hip arthroplasty in patients with nondisplaced femoral neck fractures: short-term results from a geriatric trauma registry

We performed a retrospective register-based observational study with datasets from the Registry for Geriatric Trauma of the German Trauma Society (ATR-DGU) including geriatric trauma patients with hip fractures admitted to 74 hospitals between 1 January 2016 and 31 December 2018 (n = 16,236). After the exclusion of all non-FNF patients (n = 9394), displaced FNF patients (n = 4803) and patients who had undergone treatments other than IF or HA (n = 312), the final cohort of nondisplaced FNF patients (Garden Types I and II fractures; n = 1727) included only patients treated with either IF (n = 449) or HA (n = 1278, Fig. 1). Data from the ATR-DGU comprise pseudonymized and standardized documentation of geriatric patients (≥ 70 years of age) with surgically treated hip fractures collected at five consecutive time points: upon hospital admission, preoperatively, intraoperatively, during the first postoperative week, and at discharge/transfer. Furthermore, on postoperative Day 120, an additional optional follow-up was conducted. Upon hospital admission, the patients were assessed once with the modified and validated Identification of Seniors at Risk (ISAR) screening tool [18, 19]. The ISAR score indicates the risk of adverse health outcomes, including mortality, functional decline, readmission, and institutionalization, and ranges from 0 (low risk) to 6 (high risk) points. The total number of patients with prefracture anticoagulation was recorded, and consequently, in the case of intake of anticoagulants, the individual anticoagulant medication was reported. Briefly, antiplatelet drugs (APDs) include acetylsalicylic acid and P2Y₁₂ receptor blockers, vitamin K antagonists (VKAs) include phenprocoumon and warfarin, and direct oral anticoagulants (DOACs) include factor IIa inhibitors and factor Xa inhibitors. All patients provided written informed consent for participation in the registry (ATR-DGU) after receiving spoken and written information. The present study complies with the RECORD statement [20] and is in line with the publication guidelines of the ATR-DGU [21] registered under project ID 2019–008.

Primary outcomes included mobility and residential status, reoperation rate, and HRQoL. The prefracture mobility status was categorized as ambulation without impairment or impaired ambulation depending on whether assistive devices were needed to support walking. The ambulation without impairment category included patients with the ability to walk without assistive devices. The category of impaired ambulation included patients who used walking sticks, crutches, a walker or any other mobility device either inside or outside the residency or who had no functional walking ability (i.e., no possible use of lower extremities). Furthermore, the residential status of the patients was documented and categorized as not institutionalized (community-dwelling) or institutionalized (residents of nursing homes and hospitals) for further analysis. A reoperation was defined as any surgical procedure performed due to complications after the primary (index) surgery—either during in-hospital care or during the first 120 days after the index surgery. HRQoL was measured by the non-disease-specific European Quality of Life-5 Dimensions (EQ-5D) instrument [22]. The EuroQol Group developed the standardized EQ-5D questionnaire, which expresses each respondent’s health status through a one-dimensional measure ranging from 0 (very poor) to 1 (the best possible health) [23]. The survey datasets available in this study were based on the original three-level version of the EQ-5D questionnaire (EQ-5D-3L), hereafter referred to as the EQ-5D questionnaire. Responses to the five questions of the EQ-5D questionnaire were transformed into a single HRQoL value using the time trade-off algorithm, which was validated for use in Germany [24]. The baseline EQ-5D questionnaire was administered during the first week of hospital admission and repeated prospectively during a routine follow-up of 120 days after surgery either in the outpatient clinic or via telephone interview. Mortality, both during in-hospital care and 120 days after surgery, was the secondary outcome.

The baseline characteristics and outcome variables of the study population are provided using descriptive statistics. The data are presented as medians with interquartile ranges (IQRs) for continuous variables or percentages (%) for categorical variables. A Pearson chi-square test was used to compare categorical variables with more than five expected observations, and Fisher’s exact test was applied for categorical variables with fewer than five expected observations. A Mann–Whitney U test served to compare continuous variables. While adjusting for the key covariates of sex, age, prefracture residential and mobility status, ISAR scores, intake of prefracture anticoagulation, and ASA class (American Society of Anesthesiologists classification), multivariable logistic or linear regression analyses were applied to assess the associations of surgical treatment with in-hospital mortality and EQ-5D index scores as well as mortality, reoperation rate, residential and mobility status, and EQ-5D index scores 120 days after surgery. Odds ratios (ORs) or regression coefficients (ẞ) are presented with their respective 95% confidence intervals (CIs). Two-tailed p values < 0.05 were considered significant. All analyses were performed using R statistical software (version 4.0.2; R Foundation for Statistical Computing, Vienna, Austria).

Patients who received IF were more frequently able to ambulate without any impairment at 120 days after surgery than patients in the HA group (19.6% vs. 10.5%, p = 0.003). The number of institutionalized patients at follow-up was lower in the IF group (29.3% vs. 38.2%, p = 0.039). We found no difference in terms of reoperation rates during in-hospital care (IF 2.5%, HA 2.6%, p = 1.000) or during the 120-day follow-up period (IF 5.9%, HA 4.0%, p = 0.310). Notably, however, in the IF group, requiring secondary THR or hemiarthroplasty was the major cause for reoperation during in-hospital care (3 out of 12 events) and during 120 days of follow-up (9 out of 17 events). The types of reoperation are depicted in Supplement 1 of the supplementary material. The median EQ-5D index score on day seven after surgery was 0.70 in both groups (IF 0.29–0.79, HA 0.29–0.70, p = 0.769). At the 120-day follow-up, the EQ-5D index score of the IF group was 0.81 (0.5–1.0) and that of the HA group was 0.80 (0.5–0.9, p = 0.095). Mortality rates during in-hospital care (IF 3.6%, HA 3.7%, p = 1.000) and during the 120-day follow-up period were similar between the groups (IF 7.5%, HA 9.7%, p = 0.468). The primary and secondary outcomes are summarized in Table 2.

The prognostic value of surgical treatment (IF or HA) for primary and secondary outcomes adjusted for patient demographics and baseline characteristics was analyzed in multiple individual linear or logistic regression analyses. After adjusting for key covariates, including prefracture ambulatory capacity, HA was a significant independent predictor of reduced mobility at the 120-day follow-up point (OR 2.28, 95% CI 1.11–4.74). Thus, the mobility of patients treated with HA at follow-up was 2.28 times more often impaired than that of those who received IF. Furthermore, no independent association of surgical treatment with institutionalized living status, reoperation rate, or mortality was observed (Fig. 2). Additionally, no independent associations of surgical treatment with the EQ-5D Index Score at day seven during in-hospital care (data available for 827 patients) or at 120 days after index surgery (data available for 296 patients) were found (ß = 0.010, 95% CI − 0.030–0.050, p = 0.615 and ß = 0.004, 95% CI − 0.057–0.064, p = 0.903, respectively).