*3.3. Influence on Lack of Recovery by 6th Month of the Category of the Walking Ability Prior to Admission*

3.3.1. Regarding the Previous Situation or Admission

Using Pearson's χ<sup>2</sup> tests and likelihood ratio (χRV 2) with the dependent variable recovering (yes/no) walking ability (Table 5), risk factors have been found for non-recovery at the sixth month, age ≥ 85 years, extracapsular type of fracture, ASA III or IV surgical risk, BI < 90 prior to admission, moderate or severe cognitive impairment (PS ≥ 5), institutionalization prior to admission, comorbidities at admission: chronic anemia, heart failure, having the patient prescribed anticoagulants and proton-pump inhibitors before the fracture. The poor ability to walk before admission has been significant as a protective factor for non-recovery at six months; this effect does not change and gains greater associative strength in the multivariate adjustment. There is no association with sex or with BI at admission < 60 points.

**Table 5.** Bivariate analysis recovery of the FAC (yes/no); variables significantly associated; situation prior to admission.


ASA: American Society of Anesthesiologists Physical Status Classification; BI: Barthel Index; PS: Pfeiffer Scale; FAC: functional ambulation classification.

The binary logistic regression (Table 6) has obtained the only significant results (R2 = 0.500) for not recovering walking ability, thanks to the adjustment, in addition to older age, the extracapsular type of fracture, surgical risk, number of errors in PS, and use of proton-pump inhibitor, on or prior to admission. IB (0–100) independence and, above all, worse ambulation (FAC) prior to admission have been protective factors for the lack of recovery of gait; these two effects are found in the bivariate analysis.

**Table 6.** Binary logistic regression of recovery of the FAC (yes/no); situation prior to admission.



ASA: American Society of Anesthesiologists Physical Status Classification; BI: Barthel Index.

#### 3.3.2. Regarding the Effect That the Fracture and Admission Exert

In the bivariate analysis with variables during admission and at the end of it (Table 7), we have found that the following are risk factors for non-recovery of gait: surgical technique by synthesis, start of standing, and gait beyond the third postoperative day, BI < 90 at discharge, BI < 60 at discharge, poor walking ability at hospital discharge, impairment during admission of at least one category in ambulatory ability, cognitive impairment according to PS ≥ 5, loss during admission of at least one category according to the same PS, and new institution at discharge.

**Table 7.** Bivariate analysis recovery of the FAC (yes/no); variables significantly associated; effect of the fracture and outcome.


BI: Barthel Index; FAC: Functional Ambulation Classification; RO: Odds Ratio; PI: Post Intervention.

We have also found various complications that occurred during admission as risk factors: hemoglobinemia ≤ 8.5 mg/dL, being transfused and if performed with three or more packed red blood cells, delirium, constipation, impaired kidney function during admission, urinary tract infection (UTI), acute urine retention (AUR), need to a new prescription of vitamin D at discharge.

There are four variables: deep venous thrombosis (DVT), acute ischemic stroke (AIS) during admission, liquid thickeners, and new neuroleptics prescription at hospital discharge, which are risk factors in the analysis, but the result must be interpreted with reservation, because in the 2 × 2 table, at least one box has expected values less than 5, and therefore, despite their significance, we will not include them in the multivariate analysis.

Neither the type of treatment nor the hospital stay, nor the surgical delay influence the non-recovery of walking capacity.

Institutionalization as a residential destination (new and not new) at discharge is a protector factor in the non-recovery of walking, contrary to new institutionalization, so patients with an institutional destination at discharge have significantly greater possibilities to maintain their previous level of capacity for ambulation.

The multivariate analysis (Table 8), with binary logistic regression (R<sup>2</sup> = 0.275), only confirms as true factors associated with not recovering the ability to walk the loss of at least one category of ability to walk during admission and synthesis as a technique surgery used.

**Income Events and Effects No Recovery of Baseline Walking Ability at 6th Month R<sup>2</sup> = 0.275 Coef.** β χ**<sup>2</sup> Wald** *p* **Value RO L. Inf L. Sup** Age (years) 0.075 18.318 0.000 **1.078** 1.042 1.116 Sex male 0.239 0.937 0.333 1.269 0.783 2.058 Surgical technique synthesis 0.475 4.960 0.026 **1.609** 1.059 2.445 BI at discharge (0–100) −0.004 0.218 0.641 0.996 0.981 1.012 Walking ability at discharge (FAC 1–3) −0.183 0.390 0.532 0.832 0.468 1.480 Loss walking ability during admission 0.868 14.706 0.000 **2.382** 1.529 3.712 Loss ≥ 1 category PS during admission 0.219 0.133 0.715 1.245 0.383 4.047 PS (number of errors) at discharge 0.095 2.846 0.092 1.100 0.985 1.229 Residential destination when discharged −0.433 2.694 0.101 0.648 0.386 1.088 New institutionalization at discharge 0.481 2.150 0.143 1.618 0.851 3.077 Anemia on admission 0.214 0.591 0.442 1.239 0.718 2.137 Be transfused during admission −0.171 0.435 0.510 0.843 0.506 1.402 Delirium 0.205 0.735 0.391 1.228 0.768 1.964 Constipation 0.266 1.387 0.239 1.304 0.838 2.029 Impaired renal function −0.075 0.095 0.758 0.928 0.575 1.497 UTI 0.347 1.105 0.293 1.415 0.741 2.701 AUR 0.259 0.526 0.468 1.295 0.644 2.605 New thickeners at hospital discharge 1.673 2.028 0.154 5.328 0.533 53.270 New vitamin D prescription at discharge 0.066 0.088 0.767 1.068 0.692 1.647

**Table 8.** Binary logistic regression of recovery of the FAC (yes/no); situation during admission.

BI: Barthel Index; FAC: functional ambulation classification; PS: Pfeiffer Scale; UTI: urinary tract infection; AUR: acute urine retention.

Functional loss during admission (see Tables 9 and 10, as well as Figure 2) after hip fracture in the elderly in our series is basically related to cognitive impairment before said admission, but in a different direction. There is a direct relationship or risk factor regarding the deterioration of the ability to walk. On the other hand, there is an indirect relationship so that patients with greater cognitive impairment at admission experience less loss of independence during admission.

**Table 9.** Profile of patient losing independence in at least one BI category during admission according to binary logistic regression.


BI: Barthel Index; PS: Pfeiffer Scale; FAC: functional ambulation classification.

**Table 10.** Profile of patient losing independence in at least one FAC category (≥2 levels) during admission according to binary logistic regression.


FAC: functional ambulation classification; PS: Pfeiffer Scale; BI: Barthel Index.

**Figure 2.** Summary of the interaction between functional factors in the elderly with hip fracture in our patients.

Below (Figure 2), the relationships between functional variables are exposed so that in blue, we have those that prevent and in red, those that are risk factors for non-functional recovery in the sixth month.

#### **4. Discussion**

Age is the factor that, in almost any publication, is associated with the limitation in the recovery of the previous function after a hip fracture in the elderly and in any period of time: 2 and 6 months [26]; 4 months [27], 6 months [28,29], 8 months [30], 1 year [31,32], 6 y 18 months [33], or not specifying a certain time, but when a more or less specific rehabilitation program ends [34–42]. In general, men have the worst evolution, according to much of the literature consulted [36,37,43]. According to Sylliaas et al. [44], women have a worse evolution, although there are also authors who, coincidentally with our work, do not appreciate differences [32,45]. In our study, as in the literature consulted, age, both in bivariate and multivariate analysis, is a risk factor for the non-recovery of independence and, also, for the non-recovery of ambulatory capacity.

In our investigation, the average stay is not associated with a lack of functional recovery. Martin-Martin et al. [40] associate it with worse mobility and Orive et al. [33] with BI impairment. The surgical delay in this work does not condition the functional evolution either, but there are studies in which surgical delay ≥ 48 h limits mobility [28] or the recovery of independence [33].

The pathology associated with the patient who is admitted to be treated for a hip fracture has different importance. The frailty of the elderly can be defined by the number of severe or terminal chronic diseases that the patient has [46], obtaining an index that is adjusted for age and baseline functional status. Kua J. et al. [47] have highlighted that the previously known geriatric scale [48] called Reported Edmonton Frail Scale, has a high prognostic value in all hospital admissions for acute processes in the elderly, and specifically a significant impairment (OR = 6.19, *p* = 0.01) of basic activities of daily living (ADL) [49,50] in the sixth month after hip fracture.

The number of concurrent comorbidities has been described as a factor of poor functional prognosis at four months [41] that we have not found. In fact, in our multivariate adjustment, no comorbidity influences the recovery of function at six months. Parkinson's disease has a proven relationship with ambulatory capacity in patients with hip fractures [51]. In addition, it has been described that hypertension and diabetes are comorbidities associated with a greater limitation of functional recovery [36,41]. In addition, it has been described that hypertension and diabetes are comorbidities associated with greater limitation of functional recovery [52,53]. The greater surgical risk of our patients limits the recovery of [43] independence in terms of the BI value, not as well as the recovery of the march in our research, as other authors refer [30,33].

Several authors [51,54] associate the need for help to walk or not being able to walk alone outside the residential setting before admission with not regaining independence (IADL) [55] a year after the fracture. McGilton et al. [56] consider that poor global functional status, gait, and cognitive status at admission are limiting to recovery. Lower BI and more errors in the PS impair both the global functional status and the ability to walk Mariconda M. et al. [57] at one year. In our series, cognitive impairment prior to admission limits the recovery of both independence and gait in the sixth month after the hip fracture. The most independent patients, according to the BI before the fracture in this series, are the ones with the most limited global functional recovery (BI) in the sixth month. This phenomenon and with the same index is described in the literature [33] with prospective research at 6 and 18 months. However, in our patients, functional deterioration during admission is directly related to said previous cognitive deterioration only in the case of walking. Patients with a worse baseline cognitive situation acquire a lesser loss of their independence between admission and discharge. Similarly, patients with worse gait have at admission (higher value of the FAC variable), as occurs with dependency, with less functional reserve at admission, less loss generated by the fracture, and they maintain

levels at the sixth month not as different from the previous ones. Therefore, the high value of the FAC variable prevents the non-recovery of the gait function. The essential factor so that these functions, independence and ability to walk, are not recovered is their qualitative loss during admission, especially in the case of loss of dependency (OR = 25.43, 95% CI: 12.61–51.28). Our work coincides with Dubljanin-Raspopovi´c E. et al. [27] in that cognitive impairment is a pre-eminent factor in global functional (BI) and gait non-recovery.

Our patients from a nursing home before fracture have, after adjusting variables, a recovery of BI and gait not significantly different from those who lived at home, coinciding with Ariza-Vega P. et al. [31]. Other works instead [32,42] consider that institutionalization prior to admission limits gait recovery.

The extra-articular fracture type has, in general, a worse functional prognosis in the literature [32,40,45,52], just as we have clearly found in our multivariate analysis regarding the non-recovery of gait function. The worse prognosis in the evolution of BI can, at least in part, be explained by age since our patients with extra-articular fractures have a higher mean age, as in almost all the literature [16,53]. Di Monaco [58] does not find differences in prognosis between the types of fracture. A meta-analysis [59] showed that the use of total arthroplasty in patients with displaced intracapsular fractures gives better functional results than osteosynthesis, and total hip arthroplasty, according to prospective studies, is preferable in this type of fracture both due to its functional outcome as having fewer complications [60–62]. The synthesis, in our research, by the bivariate analysis, is followed by less recovery of both dependency (BI) and walking capacity at six months. This effect, in the multivariate analysis, is annulled in terms of non-recovery of BI; and persists as a risk factor in the non-recovery of walking. The mean age of our survivors does not differ significantly between those who underwent synthesis or arthroplasty. The only complications that we have been able to relate to the functional prognosis after multivariate adjustment have been anemia, coinciding with Foss N.B. et al. [10], and constipation; however, for other authors [63], they are ulcers by pressure and "delirium".

It is a relative limitation that the measurement of the evolution at six months is a shorter time than that of some publications, which was already mentioned that they take 12 or 18 months, although there is no lack of medium-term studies: six months like ours, even at two, and four months in some cases. It has been pointed out that most of the recovery of global independence (BI) occurs in the first trimester [51]. The scientific evidence of a retrospective observational study is less than that of a cohort study, fundamentally because it is a mere consultation of registered data, no matter how rigorous the anamnesis and record of it have been. Our hip fractures do not follow any rehabilitation program, which may be related to the high percentages of lack of functional recovery that we have; in agreement with Orive et al. [33] when they state that not referring to rehabilitation increases the possibility of deterioration of the BI prior to six months, more than two times (OR = 2.34, 95% CI: 1.31–4.16) and at 18 months more than three (OR = 3.18, 95% CI: 1.62–6.25) with respect to undergoing rehabilitation treatment.

As strengths, it should be noted that the sample is large enough. Includes all fractures treated by our hospital in relation to its health area. This minimizes potential selection biases that often accompany a retrospective study. Take all possible variables. In addition to performing statistical analysis comparing dichotomous qualitative variables, binary logistic regression, in which we also incorporate quantitative independent variables for adjustment, allows us to eliminate biases such as effect modification or interaction, especially in relation to age. Although retrospective, it is still a longitudinal study, which to a large extent allows its conclusions to be taken as a valid explanation of the knowledge of the factors that truly influence limiting functional recovery in the elderly with hip fractures in our environment.

The results of this research show the factors in our population of patients aged ≥65 years, which limit, and to what extent, the recovery of the situation of independence (IB), as well as their ability to walk before suffering a hip fracture, as established in the objective of the research.

#### **5. Conclusions**

The factors associated with both the lower recovery of the BI and the ability to walk are older age and worse cognitive status at admission. Perhaps the lack of referral to rehabilitation of our patients is a very important factor to take into account in the poor recovery from dependency and walking.

Limitations to the recovery of independence are one's own independence (high BI) on admission and discharge, the loss of it during admission, and the high surgical risk (ASA).

Both dependency (low BI) as well as impaired ambulatory capacity during admission limit recovery of gait.

Patients suffering from extracapsular fractures and surgical treatment by synthesis limit the recovery of walking in the sixth month. Likewise, patients taking proton-pump inhibitors prior to admission have less recovery from walking.

Hemoglobinemia < 8.5 mg/dL, as well as constipation, are the complications that are associated with a worse prognosis of dependence, but not "delirium".

Sex does not influence, neither have any comorbidity been found, nor the greater number of concomitant comorbid processes with hip fracture related to functional prognosis

**Author Contributions:** Conceptualization, E.G.M., E.G.G. and A.d.P.M.-R.; methodology, E.G.M., E.G.G. and J.J.G.-B.; software, E.G.M., E.G.G. and M.S.-P.; validation, E.G.M., E.G.G., A.d.P.M.-R. and J.J.G.-B.; formal analysis, E.G.M., E.G.G. and M.S.-P.; investigation, E.G.M. and E.G.G.; resources, E.G.M., E.G.G., A.d.P.M.-R. and J.J.G.-B.; data curation, E.G.M., E.G.G., A.d.P.M.-R. and J.J.G.-B.; writing—original draft preparation, E.G.G. and J.G.-S.; writing—review and editing, E.G.G., M.S.-P. and J.G.-S.; visualization, E.G.M., E.G.G., A.d.P.M.-R., J.J.G.-B., M.S.-P. and J.G.-S.; supervision, J.G.-S. and J.J.G.-B.; project administration, E.G.M., E.G.G. and J.J.G.-B. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted in accordance with the Declaration of Helsinki and approved by the Ethics Committee for Drug Research of the Health Area of Burgos and Soria (CEIm 2537, approved on 27 April 2021).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Conflicts of Interest:** The authors declare no conflict of interest. Official Defense Bulletin certifies that Enrique González Marcos belongs to the Military Corps of Health in the Scale of Officers.

#### **References**

