Osteoporotic hip fracture-Comorbidities and factors associated with in-hospital mortality in the elderly: A nine-year cohort study in Brazil.

INTRODUCTION: The aim of the study was to identify factors associated with the causes of in-hospital morbidity and mortality in an elderly Brazilian population due to osteoporotic hip fractures.
METHOD: Retrospective cohort study involving a population over 60 years of age admitted to hospital due to osteoporotic hip fractures and followed up from hospitalization to outcome (discharge or mortality) from 2010 to 2018, in a public hospital in Brasília, the capital of Brazil. Multivariate analysis was performed using the Poisson regression model with a robust variance, observing the hierarchical model proposed and the receiver operating characteristic (ROC) curve to obtain the cutoff point for mortality incidence in relation the total length of hospital stay. Significance level was set as p < 0.05. The analyses were conducted using the SAS 9.4 software. RESULT: The mean hospital mortality rate among the 402 patients involved was 18.4%, and the associations made with the outcome mortality were per relevance: respiratory infection, age over 90 years, high preoperative cardiovascular risk, chronic obstructive pulmonary disease (COPD) as comorbidity, serum hemoglobin level ≤ 10 and other infections. Mortality also showed association with longer total length of hospital stay, as well as with prolonged postoperative period.
CONCLUSION: Hip fractures in the elderly due to osteoporosis indicate a relationship between the sicker profile of the aging elderly population and the prevalence of chronic diseases strongly associated with in-hospital infections, contributing to increased mortality. There were fewer early interventions, and mortality was also associated with prolonged postoperative period. The aim of this study was not to compare independent variables with each other, but suggests the relationship between the presence of comorbidities, which predisposes to the development of infections, directly linked to mortality.

Introduction

Hip fractures due to a fall from standing height are related to bone fragility and can be used to diagnose osteoporosis [1]. The factors associated with falls and risk of fractures and with the process of osteoporosis onset indicate a relationship between longevity and chronic diseases [2].

A study using data from the Brazilian National Health Survey (PNS) conducted in 2013 in Brazil revealed that the prevalence of three simultaneous diseases in the age group of 60 years or more was 3.7-fold higher than in the aged group 35 to 59 years and almost 20-fold higher than in those between 18 and 34 years old [3].

Recent studies have shown that isolated comorbidities are a dominant predictor of mortality [4, 5], being strongly associated with in-hospital infections [6]. In turn, infection is associated with a substantially increased mortality risk [7].

The surgical decision-making process for hip fracture repair in the elderly is not straightforward. The surgeon can use several tools to determine the mortality risk and define who would be more benefitted from surgical interventions and who should be referred for non-surgical interventions [8].

Mortality prediction models, which describe the outcomes distribution among the population with a particular set of characteristics, can support physicians in adapting treatments for decision making in frail elderly patients, as can causal effect estimates, which help us understand the impact of different treatment decisions made in that population [9, 10].

The aim of this study was to identify factors associated with in-hospital mortality among patients with osteoporotic hip fractures undergoing surgery or conservative treatment considering an elderly population with multiple comorbidities treated in Brazil, a developing country undergoing a rapid population aging process. This is the first Brazilian study to consider the risk attributable to the identified factors.

Methods

Data sources and configuration

Retrospective cohort study involving patients with 60 years of age or more admitted to hospital due to hip fracture after a fall from standing height and followed up from hospitalization to outcome (discharge or mortality) from 2010 to 2018.

The study was conducted in a reference public hospital in the treatment of orthopedic trauma and accredited by the Brazilian Society of Orthopedics and Traumatology (SBOT) located in Brasília, the capital of Brazil. The hospital has protocols in place for prophylaxis against deep-vein thrombosis [11].

Data were collected from medical records of patients during the hospital stay evolution and complementary examinations. Exclusion criteria were pathological fractures or undefined minimal trauma fractures.

For identification and description of the variables, the data were collected through the patients’ admission records in the Orthopedics ward in the mentioned period (2010–2018) through Trackcare@ electronic medical records. After that, each clinical record related to the identified patient was analyzed individually and sequentially in the daily record of the nursing team for data conference, where the selection of the variables of interest to the study and transcription to a database on Microsoft Excel for further statistical analysis (S1 Database and S1 File).

The mortality outcome was considered a dependent variable. The independent variables collected were (a) demographic: sex, age (60 to 79, 80 to 89, > 90 years); (b) factors associated with clinical conditions and comorbidities: hemoglobin (≤ 10 g/dL, > 10 g/dL), systemic arterial hypertension (SAH) (no, yes), diabetes mellitus (DM) (no, yes), neurological disorders (no, yes), chronic obstructive pulmonary disease (COPD) (no, yes), surgical risk (low to moderate, high), femur fracture (intracapsular, extracapsular); and (c) in-hospital variables: type of surgery (did not undergo surgery, osteosynthesis, arthroplasty), respiratory infection (no, yes), urinary tract infection (no, yes), another type of infection (no, yes) and (none, one, two or more), pulmonary thromboembolism (PTE) (no, yes), intensive care unit (ICU) stay in days (≤ 3, > 3).

An anatomical subdivision into intracapsular fracture (fracture of head and neck of the femur–S72.0) and extracapsular fracture (pertrochanteric fracture–S72.1 and subtrochanteric fracture–S72.2) was used to classify the hip fracture areas according to the International Classification of Diseases, Tenth Revision (ICD-10).

In regard to the identification of comorbidities, the selection was based on the etiological or topographic diagnosis in order to better identify the nature of the condition and its association with the mortality outcome instead of a classification score. SAH and DM were independently identified by standardized classifications [12, 13]. Concerning the infections identified during hospital stay, sites of involvement such as the lungs, urinary tract, and others were determined.

The use of the Detsky Modified Cardiac Risk Index (1986) was standardized by stratifying the patients into three risk groups (Class 1: 0–15 points (low risk); Class 2: 20–30 points (moderate risk); and Class 3: > 30 points (high risk)) [14].

Due to the varied clinical conditions, the Wells score, which considers the main risk factors without using complementary examinations, was used to estimate the clinical probability of PTE. The final sum provides an approximate patient classification as high (> 7 points), medium (2–6 points), or low (0–1 point) [15-18].

To prevent biases such as information bias, the researchers trained the technicians responsible for collecting data and controlling database entries; also, a periodical analysis and ongoing review for data inconsistency were carried out throughout the study period.

The study was approved by the Research Ethics Committee under number CAAE: 89658718.8.0000.5553, and the informed consent form was waived upon approval.

Statistical analysis

The statistical analysis was initiated with the descriptive analysis of the variables’ frequencies and determination of the incidences associated with mortality, with the respective confidence intervals (CI).

To test the effect of the independent variables on mortality, a multivariate analysis was conducted using the Poisson regression model with robust variance, observing the following hierarchical model: demographic variables composed the first stage of analysis, preoperative variables composed the second stage, and in-hospital variables composed the third stage. In this analysis, variables with a p < 0.10 were considered adjustment factors for subsequent blocks within each hierarchical level.

The Poisson regression model with robust variance was chosen because it provides a better estimate of incidence ratios, which represent the measures of effect for prospective studies, such as the relative risk (RR), in a more significant manner. The analysis took place in two phases, bivariate and hierarchical multiple regression analysis, and their RRs and respective 95% CI were calculated [19, 20].

Tolerance indicator values of less than 0.403 were the limit of multicollinearity set among independent variables.

For the quantitative time variables, the Shapiro-Wilk test was used to test the data normality, while for bivariate analysis, the non-parametric Mann-Whitney test was used to compare the groups (mortality and survival outcomes).

The receiver operating characteristic (ROC) curve was calculated to obtain the prevailing cutoff point for mortality in relation to the total length of hospital stay and to the postoperative period. The cutoff point was obtained by combining two criteria: the first, defined as the shortest Euclidean distance between the binary classification result estimated by the test and the point that provides the perfect predictor (100% sensitivity and 100% specificity); and the second, defined as the maximum Euclidean distance between the binary classification result estimated by the test and the point that provides a non-informative predictor (45° straight line).

P-values of p < 0.05 were considered significant. The analyses were conducted using the SAS 9.4 software.

Results

Four hundred and two patients who met the selection criteria were included, and the mean mortality rate as an in-hospital outcome for the period was 18.4%. The total number of elderly patients admitted to hospital with hip fractures due to falls, the distribution of the mean in-hospital mortality rate, and the mean length of hospital stay in days, per year, are described in Table 1.

Table 1

Total patients with hip fracture admitted, mean in-hospital mortality rate and mean length of hospital per year.

Admission year	2010	2011	2012	2013	2014	2015	2016	2017	2018	Total
Patients admitted (n)	45	60	40	60	51	54	31	36	25	402
In-hospital mortality	9	8	8	11	13	12	5	4	4	74
Mean in-hospital mortality rate (n, %)	20.0%	13.3%	20.0%	18.3%	25.5%	22.2%	16.1%	11.1%	16.0%	18.4%
Mean length of hospital stay (days)	23.4	20.1	28.2	21.6	27.2	30.0	27.5	28.4	25.5	25.4

The descriptive analysis is described in Table 2, showing the characteristics of patients sustaining fractures and mortality according to the studied variables. As for the demographic variables, a predominance of fractures in women (n = 260; 64.68%), but higher mortality in men (20.42%; 95% CI 13.76–27.08), was found. The age group 80–89 years had a higher incidence of fractures (n = 142; 32.34%); however, the highest proportional mortality was in the age group over 90 years (52%; 95% CI 30.67–57.79).

Table 2

Characteristic of patients and mortality.

Variables		Patients (n = 402)	Porcentage (%)	Mortality (%)	95% IC
Sex	Women	260	64.68	17.31	12.69–21.93
	Men	142	35.32	20.42	13.76–27.08
Age	60–69	78	19.40	5.13	0.21–10.04
	70–79	130	32.34	8.46	3.66–13.27
	80–89	142	35.32	25.35	18.17–32.54
	> 90	52	12.94	44.23	30.67–57.79
Hemoglobin	≤ 10	87	21.64	28.74	19.19–38.28
	> 10	315	78.36	15.56	11.54–19.57
Comorbidities	No	49	12.19	8.16	0.46–15.86
	Yes	353	87.81	19.83	15.65–24.01
SAH	No	97	24.13	12.37	5.79–18.95
	Yes	305	75.87	20.33	15.79–24.86
DM	No	263	65.42	16.35	11.86–20.85
	Yes	139	34.58	22.30	15.35–29.25
Neurological disorders	No	296	73.63	15.88	11.70–20.06
	Yes	106	26.37	25.47	17.14–33.80
COPD	No	371	92.29	15.90	12.17–19.64
	Yes	31	7.71	48.39	30.72–66.05
Surgical risk	Low to moderate	268	66.67	8.58	5.21–11.95
	High	134	33.33	38.06	29.80–46.32
Femur fracture	Intracapsular	218	54.23	18.81	13.60–24.02
	Extracapsular	184	45.77	17.93	12.37–23.50
Type of surgery	Non-surgical	92	22.89	38.04	28.08–48.01
	Osteosynthesis	173	43.03	13.87	8.70–19.05
	Arthroplasty	137	34.08	10.95	5.70–16.20
Infection	No	267	66.42	2.25	0.46–4.03
	Yes	135	33.58	50.37	41.90–58.84
Respiratory infection	No	316	66.42	4.43	2.15–6.71
	Yes	86	21,39	69.77	60.02–79.51
Urinary tract infection	No	338	84.08	14.20	10.46–17.94
	Yes	64	15.92	40.62	28.54–52.71
Other infection	No	386	96.02	17.36	13.56–21.15
	Yes	16	3.98	43.75	19.34–68.16
PTE	No	384	95.52	16.41	12.69–20.13
	Yes	18	4.48	61.11	38.48–83.73
Days in ICU	≤ 3	322	80.10	12.11	8.53–15.69
	> 3	80	19.90	43.75	32.83–54.65

Regarding the preoperative assessment variables, most patients had a low to moderate cardiac risk (n = 268; 66%), although a high number of patients (n = 353; 87.81%) presented with some comorbidity at the time of hospital admission. SAH (n = 305; 75.87%) and DM (n = 139; 34.58%) were the most frequent comorbidities. Almost half of the patients (n = 218; 54.23%) also had other comorbidities concomitant with those selected for analysis, such as chronic kidney disease (6%), hypothyroidism (8%), and dyspeptic disorders (5%). As for the type of fractures, the intracapsular (femoral neck) ones presented a slight predominance (n = 218; 54.23%).

Considering the other variables resulting from the hospital stay, most patients (n = 310; 77.11%) underwent surgical procedures, and the most frequently performed one was osteosynthesis (n = 173; 13.87%). Nonetheless, the highest mortality was found among patients who did not undergo surgery (n = 92; 38.04%). Most patients did not present infection during hospital stay (n = 267; 66%) and stayed less than three days in the ICU (n = 322; 80,1%). The number of patients diagnosed with PTE was low (n = 18; 4.48%).

Despite this, the highest incidence of mortality occurred among patients who had some type of infection (n = 135; 34%). Considering the affected topographic sites, 86 patients with respiratory infection (21.39%), the outcome death occurred in 69.77% (60.02–79.5), followed by those with urinary infection (n. = 64; 15.92%), with 40.62% (28.54–52.71) mortality. The same occurred with patients who remained in the ICU for more than 3 days (43.75%; 32.83–54.65) and had pulmonary embolism (n = 18; 38.48–83.73).

Mortality and risk variables

In the bivariate analysis, a statistically significant association was observed between mortality and the following variables: age groups 80–89 years (RR = 3.52; 95% CI 2.00–6.17) and over 90 years (RR = 6.13; 95% CI 3.45–10.90); hemoglobin ≤ 10 (RR = 1.85; 95% CI 1.21–2.81); neurological disorders (RR = 1.60; 9% CI 1.06–2.44), COPD (RR = 3.04; 95% CI 1.97–4.69), and high surgical risk (RR = 4.43; 95% CI 2.84–6.93).

Among the in-hospital variables, non-surgical interventions (RR = 3.47; 95% CI 2.01–5.99); respiratory infection (RR = 15.75; 95% CI 9.26–26.77), urinary tract infection (RR = 2.86; 95% CI 1.93–4.25), or other infections (RR = 2.52; 95% CI 1.39–4.58); PTE (RR = 3.72; 95% CI 2.42–5.74); and a stay in ICU for more than three days (RR = 3.61; 95% CI 2.46–5.31) were also significantly associated with the mortality outcome (Table 3).

Table 3

Relative risk and prevalence ratio using the Poisson regression model with robust variance and its respective 95% confidence interval.

Variables	Relative Risk (RR)		Relative Risk (RR) Adjusted*
	RR (95% IC)	p-value	RR (95% IC)	p-value
Block 1 –Demographic
Sex		0.4394		0.5128
Women	1	-	1	-
Men	1.18 (0.78–1.79)	0.4394	1.14 (0.77–1.69)	0.5128
Age		< 0.0001		< 00001
60–79	1	-	1	-
80–89	3.52 (2.00–6.17)	< 0.0001	350 (1.99–6.15)	< 0.0001
> 90	6.13 (3.45–10.90)	< 0.0001	6.11 (3.44–10.87)	< 0.0001
Block 2 –Preoperative
Hemoglobin				0.0334
≤ 10	1.85 (1.21–2.81)	0.0041	1.54 (1.03–2.29)	0.0334
> 10	1	-	1	-
SAH		0.0901		0.2475
No	1	-	1	-
Yes	1.64 (0.92–2.92)	0.0901	1.39 (0.79–2.45)	0.2475
DM		0.1412		0.6915
No	1	-	1	-
Yes	1.36 (0.90–2.06)	0.1412	0.92 (0.63–1.36)	0.6915
Neurological disorders		0.0267		0.5050
No	1	-	1	-
Yes	1.60 (1.06–2.44)	0.0267	1.14 (0.77–1.69)	0.5050
COPD		< 0.0001		0.0002
No	1	-	1	-
Yes	3.04 (1.97–4.69)	< 0.0001	2.39 (1.52–3.78)	0.0002
Surgical risk		< 0.0001		< 0.0001
Low to moderate	1	-	1	-
High	4.43 (2.84–693)	<0.0001	3.18 (2.00–5.06)	< 0.0001
Femur fracture		0.8222		0.1681
Intracapsular	1	-	1	-
Extracapsular	0.95 (0.63–1.44)	0.8222	1.32 (0.89–1.97)	0.1681
Block 3 –Postoperative
Type of surgery		< 0.0001		0.0003
Non-surgical	3.47 (2.01–5.99)	< 0.0001	2.04 (1.31–3.16)	0.0003
Osteosynthesis	1.27 (0.69–2.32)	0.4431	1.03 (0.64–1.64)	0.9108
Arthroplasty	1	-
Respiratory infection		< 0.0001		< 0.0001
No	1	-	1	-
Yes	15.75 (9.26–26.77)	< 0.0001	7.27 (3.98–3.26)	< 0.0001
Urinary tract infection		< 0.0001		< 0.0001
No	1	-	1	-
Yes	2.86 (1.93–4.25)	< 0.0001	2.04 (1.44–2.89)	< 0.0001
Other infection		0.0024		0.0251
No	1	-	1	-
Yes	2.52 (1.39–4.58)	0.0024	1,98 (1.09–3.62)	0.0251
PTE		< 0.0001		0.0202
No	1	-	1	-
Yes	3.72 (2.42–5.74)	< 0.0001	1.98 (1.11–3.52)	0.0202
Days in ICU		< 0.0001		0.1683
≤ 3	1	-	1	-
> 3	3.61 (2.46–5.31)	< 0.0001	1.27 (0.90–1.80)	0.1683

Using the Poisson regression model with robust variance, in the first stage of the hierarchical model (epidemiological block) the sex and age variables were included. The age groups 80–89 and over 90 years showed a significant association with mortality. Therefore, the age variable was maintained for the next block analysis.

In the second stage, age was included along with the block of in-hospital variables, and only the variables hemoglobin ≤ 10 (PR = 1.54; 95% CI 1.03–2.29), COPD (PR = 2.39; 95% CI 1.52–3.78), and high surgical risk (PR = 3.18; 95% CI 2.00–5.16) showed a significant association with mortality, even after adjustment. These variables were maintained for the next block analysis.

In the last stage, in addition to the variables age, hemoglobin, COPD, and high surgical risk, the postoperative variables block was included. After adjustment for possible confounders, the variables respiratory infection (PR = 7.27; 95% CI 3.98–13.26), urinary tract infection (PR = 2.04; 95% CI 1.44–2.89), other infections (PR = 1.98; 95% CI 1.09–3.62), and PTE (PR = 1.98; 95% CI 1.11–3.52), all in relation to the bivariate analysis, showed a significant association with mortality.

Mortality and time variables

Considering the entire sample, patients who died had a longer hospital stay (31.28 ± 23.24); however, when comparing each variable, related to time with mortality outcome (yes or no), the postoperative period (21.05 ± 21.71) and longer stay in ICU (10.65 ± 16.62) presented a significant association with mortality (Table 4).

Table 4

Total mean of hospital stay, preoperative period, postoperative period and days in ICU when comparing the surviving and mortality groups.

Variables	Patients (n)	Mean (days)	SD	Mortality		p-value*
				Yes (days)	No (days)
Total mean of hospital stay (days)	402	25.39	18.91	31.28±23.24	24.06±17.56	0.0283
Preoperative period (days)	311	19.22	14.44	18.62±12.07	19.30±14.76	0.8191
Postoperative period (days)	308	7.23	10.47	21.05±21.71	5.23±5.20	<0.0001
Stay in ICU (days)	402	3.71	8.82	10.65±16.62	2.15±4.51	<0.0001

* to calculate p-value the Mann-Whitney test was used

When comparing and analyzing the time variables among the groups, the mean total length of hospital stay was significantly longer among patients who died in both groups and in both those undergoing arthroplasty (p = 0.0015) and osteosynthesis (p = 0.0005).

The same was identified for the mean postoperative period, which was significantly longer among patients who died in both groups and both among those undergoing arthroplasty (p < 0.0001) and osteosynthesis (p < 0.0001).

In turn, the mean preoperative period did not show any significant difference between the surviving and mortality groups and both in those undergoing arthroplasty (p = 0.6614) and osteosynthesis (p = 0.4211), as shown in Table 5.

Table 5

Total mean of hospital stay, preoperative period, postoperative period and days in ICU when comparing the surviving and mortality groups, regarding the surgery performed.

	Arthroplasty			Osteosynthesis			Non-surgical
Variables#	Mortality			Mortality			Mortality
	No	Yes	p-value*	No	Yes	p-value*	No	Yes	p-value*
Total mean of hospital stay	29.04±24.09	48.67±28.26	0.0015	20.82±10.54	33.29±19.52	0.0005	21.86±12.93	22.46±18.88	0.4914
Preoperative period	22.57±18.79	19.73±14.85	0.6614	16.75±9.68	17.92±10.26	0.4221	-	-	-
Postoperative period	5.91±5.86	29.27±26.57	<0.0001	4.42±3.11	15.92±16.64	<0.0001	-	-	-
Stay in ICU	2.77±5.08	18.33±21.90	<0.0001	2.24±4.29	12.88±16.78	<0.0001	0.58±3.30	5.83±12.33	0.0002

The mean stay in ICU in days was significantly longer among patients who died in all groups, both for those undergoing arthroplasty (p < 0.0001) and osteosynthesis (p < 0.0001) and for those who did not undergo surgery (p = 0.0002).

Finally, to estimate the number of postoperative days, regardless of the surgical technique employed, after which there was a statistical significance concerning the incidence of mortality in these patients, an analysis was performed using the ROC curve (Fig 1). After 22 days (95% CI 0.4998–0.6634) of hospital stay, there was a correlation with mortality in the entire sample. The same applied six days after surgery, regardless of the surgical technique employed (95% CI 0.6884–0.8959) (Figs 2 and 3) (S1 Database).

Fig 1

Process model of data extraction and patients’ follow-up during hospitalization.

Fig 2

Cutoff point for mortality from the 22nd day of the total length of hospital stay.

ROC (curve area 0.5816; 95% IC 0.4998–0.6634).

Fig 3

Cutoff point for mortality from the sixth day on for the postoperative period.

ROC (curve area 0.7922; 95% IC 0.6884–0.8959).

Discussion

Hip fractures are strongly associated with increased mortality rates in different studies worldwide [21]. In recent years, predictive models for mortality risk after hip fractures have been developed to identify patients at a higher risk and propose intervention strategies to improve the outcomes of hip fractures [9, 22–25].

The mean hospital mortality rate considering the nine years of study was 18.4%, and the variables with a significant association with the mortality outcome in elderly patients hospitalized for osteoporotic hip fractures were: 1) demographic factors: age over 90 years; 2) factors associated with clinical conditions and comorbidities: high preoperative cardiovascular risk, hemoglobin ≤ 10, chronic obstructive pulmonary disease (COPD); 3) hospital factors: respiratory infection, urinary tract infection, and other infections, pulmonary thromboembolism (PTE).

The challenge of establishing a pattern among the several variables associated with worsening and to mortality outcomes [26] is related to the fact that these associations, as well as the mortality rate assessments reported in different studies involving different populations and methodologies, also present regional variations [17, 18, 27–29].

In our study, by examining the variables selected for analysis, we can try to identify the sample behavior synthesizing the existing relationships. However, as the study purpose was not to compare the independent variables with each other, but to compare their association with the mortality outcome, we cannot attest, but only discuss the existence of a causal relationship among them.

When proposing this study, the authors gathered from the observation of real-life outcomes—discharge or death, the several hypotheses that originated the research questions, either from the health conditions of the population assisted, or the course of fracture treatment from admission to the intervention, and studied a method of data collection and analysis development, regarding the variables of interest, that demonstrated significance for the construction of evidence.

Considering the descriptive analysis, demographic factors show a higher incidence of hospital admission in elderly women, which is meets the profile of patients affected by osteoporotic femur fractures after a fall from standing height [1]. However, in the frequency analysis, the incidence of male mortality in the general sample was higher. Studies show that the excess mortality in men remains high for up to 20 years after the fracture [30], but the causes for this difference in mortality in absolute terms are not fully explained yet [31, 32]. A comparative study between sexes attributed the difference observed to the relationship between deaths and infections (pneumonia, influenza, and septicemia) [33].

Considering the increase in life expectancy worldwide, especially in developing countries [34], the association among longevity, osteoporotic hip fractures, and mortality becomes relevant [29], raising the question of how these injuries affect health systems [35]. The age group over 90 years, which was identified by the hierarchical model of multiple regression as a factor associated with a six-fold higher RR for death [RR 6.11 (3.44–10.87)], relates the factors attributed to the senescence of this age group [36] both with bone degeneration, which accounts for the highest incidence of fracture mechanisms [37], and with a higher risk of fracture complications [38].

Other factors in our study can also contribute to the regional population analysis and reflect the health conditions related to aging. Almost 90% of patients over 60 years of age presented with some comorbidity at hospital admission. The literature already described that the comorbidities identified at admission are related to the mortality outcome in patients with hip fracture in the short and long term [2]. We did not use quantitative comorbidity rates in our study but choose to present them as individual variables. It seems desirable to efficiently summarize one or more comorbidities in a single score [39]; however, the purpose of this work was to obtain a greater nosological perspective of each variable behavior separately in relation to mortality and their relationship with osteoporosis as an underlying disease [40, 41].

Among the comorbidities grouped as preoperative variables, the Poisson regression model with robust variance related COPD with a significant 2-fold higher mortality risk compared to non-COPD patients. A study also compared the mortality between COPD and non-COPD patients and revealed that COPD was an independent mortality factor over a minimum follow-up period of one year and that the disease severity in patients with hip fracture was also a risk factor for mortality for six months to one year [42].

Hemoglobin was also selected to analyze the impact of blood levels below 10 g/dL on the mortality outcome. Elderly patients with hip fractures are known to have a high risk of perioperative anemia due to blood loss related to the fractures and/or surgery [43]. Besides, studies have shown that patients who need blood transfusion require a longer hospital stay [44]. In our study, even after adjusting the model variables, there was a significant 1.5-fold higher risk of mortality in patients with levels above 10 g/dL, making it an important variable to be analyzed.

Regarding the variable DM, it should be highlighted that, despite not being included in the multiple regression model as a variable directly associated with the mortality risk, the group of patients with DM had higher mortality in the frequency description. Studies already related mortality after hip fractures with DM [45], which represents an increased risk factor in case of fractures due to fragility that seems to be independent of bone mineral density [46].

Preoperative cardiac risk was analyzed using the Detsky Modified Cardiac Risk Index, 1986 [47], and patients classified as high surgical risk (RR 10.6) were significantly associated with the highest risk of death in our study. This assessment can be used to estimate possible risks resulting from the surgical procedure in each patient and, if possible, conducts to minimize these risks. This estimate is essential to provide the surgeon/team and patient/family with information that must be taken into account when comparing the procedure’s possible benefits and harms in each case [48].

The association between atherosclerotic cardiovascular diseases and osteoporosis emphasizes epidemiological and physiopathogenic similarities in arterial wall calcification and osteogenesis considering the low bone mass, osteoporosis fractures, vascular calcification, extension of coronary and abdominal aorta injury, and cardiovascular mortality, regardless of age. Osteopenia and osteoporosis in the femoral neck were associated with a higher risk of severe coronary lesions [49, 50], and the reverse [50].

However, the Detsky Modified Cardiac Risk Index, 1986 [47] does not consider the SAH, present in 75% of patients in our study, in its score. Although this variable was not associated with a direct risk of death, patients with SAH had a higher mortality.

The high prevalence of cardiovascular diseases found in the elderly population in Brazil [51] raises two issues: 1) the established relationship between cardiovascular diseases and osteoporosis [52], and 2) SAH as an independent risk factor for increased mortality [53] to be considered, requiring evaluation regarding its prevention and associated risk factors.

Nonetheless, infection was the prevailing factor for in-hospital mortality in elderly patients with femoral fractures in this study, as per the hierarchical model of multiple regression after adjustment for possible confounders. Patients with pulmonary infection had a seven-fold higher risk of death than patients who did not have this condition.

Previous studies demonstrate pulmonary infection during hospital stay as a risk factor for death associated with one of the most common complications [53, 54], also representing an independent risk for early readmission after hip fracture surgery [55].

Evidence shows a relationship between trauma and an age-related decline in the elderly health, affecting the neutrophils function and reducing their immune response to bacteria [56]. Additionally, in patients who already have a higher incidence of pre-existing cardiorespiratory disease and reduced mobility after a hip fracture, there is an increased risk for pneumonia [55]. These were characteristics presented by the patients included in this study.

Other infections were also associated with the mortality risk in this population, such as urinary tract infections, which complies with findings of other studies that made this correlation with other complications similar to surgical site infection [18], such as progression to sepsis and extended hospital stay [57].

Regarding the hip fracture trauma mechanism pattern, these are classified by their site as an evidence of prognostic implications [58]. Surgery should be the most appropriate option for most patients [59], as it is associated with a shorter hospital stay and better rehabilitation [60]. Non-surgical interventions are reserved for patients with a severe debilitation, unstable patients with incurable severe diseases, or patients with terminal illnesses in the final stages of life [61]. Stable impacted fractures can also be considered for non-surgical treatment [62]. However, with conservative treatment, rehabilitation will probably be slower and limb deformity, more common [63].

Further studies are required to allow a shared decision-making, and questions about the pre-fracture quality of life and future perspectives should be asked before considering different treatment options to assess which one is advisable in frail and high-risk elderly patients, considering that most patients with hip fracture also have advanced comorbidities [8].

Finally, we analyzed the time variables with and without surgery comparing the techniques employed, which did not differ in relation to the risk of death. However, it was found that an increased total length of hospital stay, prolonged postoperative period, and longer ICU stay are associated with a higher mortality, which also applies to patients who have not undergone surgical treatment.

Analysis limitations should be considered for associations related to increased postoperative mortality, such as the performance or not of early mobilization after surgery. However, PTE occurrence and its relationship with mortality can suggest a deficiency in postoperative rehabilitation protocols.

There are also considerations regarding the intervention period as a limiting factor. In our study, the preoperative period did not significantly influence mortality, in contrast to studies that recommend early intervention [10]. However, due to fewer early intervention, the analysis in this group may have been insufficient for the effect of an association with significance.

On the other hand, chronic diseases can contribute to the hypothesis of aggravation or worsening, leading to the fall from their height, and therefore the health status at admission that does not allow early surgical intervention, as well as contributing further complications after surgical stress.

Despite the Brazilian Unified Health System being universal and of unrestricted access, health interventions may be conditioned to the availability of resources, equipment, and management issues of the public system, which is a reason to question also whether there were limitations in the management of the pathology, or if only the patient’s previous health conditions at admission, such as multiple comorbidities, made surgery unfeasible if there was no early intervention, considering the need for a period of clinical stability for the patient to be suitable for surgery.

A study on the length of hospital stay for the pathology in public hospitals in Brazil, in a 10-year historical analysis, considering 480,652 hospitalizations, confirms the longest mean length of hospital stay among Brazilian capitals, in the Federal District (18.7 days), the geographical region where the research was developed [64].

Therefore, this study showed the importance of comprehensive treatment in the fracture approach, from the clinical condition of the patients to treatment time, including rehabilitation and clinical care after the intervention. We suppose that often after discharge, complications are not reported (underreporting). The extended in-hospital observation of this study, on the other hand, was able to capture.

Conclusions

Although the number of hospitalizations was higher in elderly women with multiple comorbidities, the male mortality was higher. Among the comorbidities studied, mortality rates were higher in patients with diabetes, hypertension, COPD, and risk attributed to a high preoperative risk score and lower serum hemoglobin levels. The impact of these comorbidities, acting as "correlation variables," on the in-hospital variables should be considered, since the highest mortality occurs in patients who did not undergo surgery or who were hospitalized for longer periods, including ICU stay, therefore attributing a greater severity in the condition of these patients upon admission. Among the in-hospital variables, infections were the most prevalent factor associated with in-hospital mortality, especially respiratory ones. The aim of the study was not to compare independent variables with each other, but the association of comorbidities leading to the development of infections directly linked to mortality is clear. These factors emphasize the attention required by individual perspectives on healthy aging promotion and by politics and programs ensuring access to the health system, preventing comorbidities and falls and creating strategies for early risk assessment, and causal effect estimates, which help us understand the impact of different treatment decisions made in that population, to prevent the mortality outcome.

Faced with the evidence generated by the scientific community worldwide on mortality in the aspect of femur fracture in the elderly, the study sought dialogue between the results of this study and the international bibliographic references. The point discussed in the review regarding the low early intervention in this study was also honestly mentioned in the text.

The study aimed, within the methods’ limits, to separate the variables related to the individual and the variables related to the health system in the care of these patients, not to generate confusion but to clarify to the readers the main aspects that would be associated to the mortality outcome.

Study’s database.

(XLSX)

Click here for additional data file.

Statistical reporting.

(DOCX)

Click here for additional data file.

22 Mar 2022

A rebuttal letter that responds to each point raised by the academic editor and reviewer(s). You should upload this letter as a separate file labeled 'Response to Reviewers'.

A marked-up copy of your manuscript that highlights changes made to the original version. You should upload this as a separate file labeled 'Revised Manuscript with Track Changes'.

An unmarked version of your revised paper without tracked changes. You should upload this as a separate file labeled 'Manuscript'.

If you would like to make changes to your financial disclosure, please include your updated statement in your cover letter. Guidelines for resubmitting your figure files are available below the reviewer comments at the end of this letter.

If applicable, we recommend that you deposit your laboratory protocols in protocols.io to enhance the reproducibility of your results. Protocols.io assigns your protocol its own identifier (DOI) so that it can be cited independently in the future. For instructions see: https://journals.plos.org/plosone/s/submission-guidelines#loc-laboratory-protocols. Additionally, PLOS ONE offers an option for publishing peer-reviewed Lab Protocol articles, which describe protocols hosted on protocols.io. Read more information on sharing protocols at https://plos.org/protocols?utm_medium=editorial-email&utm_source=authorletters&utm_campaign=protocols.

We look forward to receiving your revised manuscript.

Kind regards,

Robert Daniel Blank, MD, PhD

Academic Editor

PLOS ONE

Journal Requirements:

When submitting your revision, we need you to address these additional requirements.

1. Please ensure that your manuscript meets PLOS ONE's style requirements, including those for file naming. The PLOS ONE style templates can be found at

https://journals.plos.org/plosone/s/file?id=wjVg/PLOSOne_formatting_sample_main_body.pdf and https://journals.plos.org/plosone/s/file?id=ba62/PLOSOne_formatting_sample_title_authors_affiliations.pdf

2. In your Data Availability statement, you have not specified where the minimal data set underlying the results described in your manuscript can be found. PLOS defines a study's minimal data set as the underlying data used to reach the conclusions drawn in the manuscript and any additional data required to replicate the reported study findings in their entirety. All PLOS journals require that the minimal data set be made fully available. For more information about our data policy, please see http://journals.plos.org/plosone/s/data-availability.

Upon re-submitting your revised manuscript, please upload your study’s minimal underlying data set as either Supporting Information files or to a stable, public repository and include the relevant URLs, DOIs, or accession numbers within your revised cover letter. For a list of acceptable repositories, please see http://journals.plos.org/plosone/s/data-availability#loc-recommended-repositories. Any potentially identifying patient information must be fully anonymized.

Important: If there are ethical or legal restrictions to sharing your data publicly, please explain these restrictions in detail. Please see our guidelines for more information on what we consider unacceptable restrictions to publicly sharing data: http://journals.plos.org/plosone/s/data-availability#loc-unacceptable-data-access-restrictions. Note that it is not acceptable for the authors to be the sole named individuals responsible for ensuring data access.

We will update your Data Availability statement to reflect the information you provide in your cover letter.

[Note: HTML markup is below. Please do not edit.]

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Partly

Reviewer #2: Yes

**********

2. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: No

**********

4. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: the authors have captured 400 hip fracture patients over 9 years and characterized the factors that influenced mortality. Their results are consistent with a number of other similar publications. This is unique in that it addressed Brazil medicine.

There are some key concerns with the patient population. This study enters less than one patient per week for hip fractures. The usual general hospital that handles hip fractures sees one per day. Were the 400 patients all the hip fractures at that hospital. If no how were they chosen and was there potential bias. If these are all the patients then the hospital is weak in hip fracture experience and why so few? Where did these patients come from?

Also of concern is the long hospital stay. Around the world most hip fracture patients leave after 4-5 days going either home or to a rehab hospital. How do the authors explain this markedly long hospital stay. Are the patients both acute and then subacute??

Time to surgery has consistently been demonstrated to influence mortality. Why not in this group?? It is recommended that the patients should undergo surgery ideally within 24 hours of admission and certainly within the first 48 hours. How did this hospital rate and did the patients who were operated within 24 hours do compared to the others?

Most of the medical issues are covered by the Charlson index and most of the OR risk is covered by the ASA score. Please provide them for this patient population

this study has a very high mortality rate. give us the 3-day rate rather than the 22-day rate.

Reviewer #2: This is a retrospective multivariate analysis from Brazil of 402 patients with hip fractures over eight years to examine factors associated with in hospital mortality. The in-hospital mortality was 18% and not surprisingly older patients, those with comorbidities, those selected for non-operative treatment, and who develop infections were at risk for higher mortality.

The study was very comprehensive and its analysis of important variables and include some relatively novel methods of stratifying risk such as cardiac, pulmonary embolism etc. with quantitative scales. The Multi variable analysis was extremely well done and eloquently explained in the results. The results and conclusions are justified by their data and analytic methods. The manuscript was extremely well written and a pleasure to read.

A number weaknesses need to be taken into consideration when interpreting those data. First I do not think the ROC analysis on length of stay is compelling or generalizable and depends on local conditions. I do not believe it is useful and would be best deleted from the study. The selection bias regarding surgical versus non-surgical treatment is problematic and clearly results in the patients treated nonoperatively having higher mortality. One approach might be the patient altogether and only discussing hospital mortality of the operative patients or do this separately. The follow up is very short and only regarding inpatient mortality which may not generalize well to the rest of the world given the prolong length of stay seen in Brazil.

101 It seems like the aims should be stated in the introduction not the conclusion. Also, in the abstract you did not comment on the results regarding infections.

145 What was the time length of follow-up? Is this only in hospital mortality?

149 you have a very good list of comorbidities however the cognitive state of the patient either acute or chronic are important predictors of outcome, risk of complications, mortality, and are perhaps the most important aspect regarding surgical decision making.

199 the ROC analysis regarding length of stay is very dependent on the cultural and healthcare system utilized in Brazil and does not apply elsewhere. Also, significant bias in the patient to die early would have short length of stay but those who have prolong illness as long as he stays would have prolonged length of stay this would seem to confound this analysis. Please explain what postoperative time period means.

305 Please explain what groups you were referring to.

324 Need to define whether this is total hip arthroplasty or hemi arthroplasty.

325 One of the major weaknesses of the study is that you do not specify indications for osteosynthesis, arthroplasty or non-operative treatment. These are largely dependent fracture type as osteosynthesis would typically use for intertrochanteric, subtrochanteric ,and some femoral neck fractures where is hemiarthroplasty for displaced femoral neck fractures. Since these cannot be used interchangeably it makes no sense comparing these surgeries. Further non operative care is typically selected for those with greater trochanteric fractures or in patients too sick to have surgery or who have a very short life expectancy. Thus, significant selection bias is present which accounts for the higher mortality.

463 I think you need to significantly expand study limitations particular problem of selection bias, unknown indications for surgery, not including cognitive state patient, only including in-hospital mortality.

**********

6. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

[NOTE: If reviewer comments were submitted as an attachment file, they will be attached to this email and accessible via the submission site. Please log into your account, locate the manuscript record, and check for the action link "View Attachments". If this link does not appear, there are no attachment files.]

While revising your submission, please upload your figure files to the Preflight Analysis and Conversion Engine (PACE) digital diagnostic tool, https://pacev2.apexcovantage.com/. PACE helps ensure that figures meet PLOS requirements. To use PACE, you must first register as a user. Registration is free. Then, login and navigate to the UPLOAD tab, where you will find detailed instructions on how to use the tool. If you encounter any issues or have any questions when using PACE, please email PLOS at figures@plos.org. Please note that Supporting Information files do not need this step.

6 May 2022

Dear Editor,

Thank you for considering the paper " Osteoporotic hip fracture – comorbidities and factors associated with in-hospital mortality in the elderly: a nine-year cohort study in Brazil" by Peterle et al., for publication in the PlosOne Journal and allowing us to re-submit this revised manuscript as a letter to the editor.

The authors are grateful to the reviewers for the careful appraisal, positive comments, and helpful criticisms. Suggested changes have been addressed. We believe that the quality of the manuscript has been improved and hope that it now meets the quality required for publication in the PlosOne Journal.

A detailed point-by-point response is given below.

We are looking forward to your decision regarding the suitability of the revised version of this paper for the Journal.

Thank you very much in advance,

Viviane Cristina Uliana Peterle*, MD, PhD

*Corresponding Author

E-mail: vivianepeterle@hotmail.com

Orcid: https://orcid.org/0000-0003-1693-242X

Afiliation 1: Escola Superior de Ciências da Saúde (Escs/Fepecs), Brasília, DF, Brazil Afiliation 2: Universidade de Brasilia, Brasília, DF, Brazil

Submitted filename: Response to reviewers.docx

Click here for additional data file.

30 May 2022

4 Jul 2022

Dear Editor,

Thank you for considering the paper "Osteoporotic hip fracture – comorbidities and factors associated with in-hospital mortality in the elderly: a nine-year cohort study in Brazil" by Peterle et al., for publication in the PlosOne Journal and allowing us to re-submit this revised manuscript as a letter to the editor.

The authors are grateful to the reviewers for the careful appraisal, positive comments, and helpful criticisms. Suggested changes have been addressed. We believe that the quality of the manuscript has been improved and hope that it now meets the quality required for publication in the PlosOne Journal.

A detailed point-by-point response is given below.

We are looking forward to your decision regarding the suitability of the revised version of this paper for the Journal.

Thank you very much in advance,

Viviane Cristina Uliana Peterle*, MD, PhD

*Corresponding Author

E-mail: vivianepeterle@hotmail.com

Orcid: https://orcid.org/0000-0003-1693-242X

Afiliation 1: Escola Superior de Ciências da Saúde (Escs/Fepecs), Brasília, DF, Brazil Afiliation 2: Universidade de Brasilia, Brasília, DF, Brazil

Submitted filename: Response to reviewers.docx

Click here for additional data file.

12 Jul 2022

Osteoporotic hip fracture – comorbidities and factors associated with in-hospital mortality in the elderly: a nine-year cohort study in Brazil

PONE-D-22-03245R2

Dear Dr. Peterle,

We’re pleased to inform you that your manuscript has been judged scientifically suitable for publication and will be formally accepted for publication once it meets all outstanding technical requirements.

Within one week, you’ll receive an e-mail detailing the required amendments. When these have been addressed, you’ll receive a formal acceptance letter and your manuscript will be scheduled for publication.

An invoice for payment will follow shortly after the formal acceptance. To ensure an efficient process, please log into Editorial Manager at http://www.editorialmanager.com/pone/, click the 'Update My Information' link at the top of the page, and double check that your user information is up-to-date. If you have any billing related questions, please contact our Author Billing department directly at authorbilling@plos.org.

If your institution or institutions have a press office, please notify them about your upcoming paper to help maximize its impact. If they’ll be preparing press materials, please inform our press team as soon as possible -- no later than 48 hours after receiving the formal acceptance. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information, please contact onepress@plos.org.

Kind regards,

Robert Daniel Blank, MD, PhD

Academic Editor

PLOS ONE

Additional Editor Comments (optional):

Reviewers' comments:

Reviewer's Responses to Questions

Comments to the Author

1. If the authors have adequately addressed your comments raised in a previous round of review and you feel that this manuscript is now acceptable for publication, you may indicate that here to bypass the “Comments to the Author” section, enter your conflict of interest statement in the “Confidential to Editor” section, and submit your "Accept" recommendation.

Reviewer #1: All comments have been addressed

Reviewer #2: All comments have been addressed

**********

2. Is the manuscript technically sound, and do the data support the conclusions?

The manuscript must describe a technically sound piece of scientific research with data that supports the conclusions. Experiments must have been conducted rigorously, with appropriate controls, replication, and sample sizes. The conclusions must be drawn appropriately based on the data presented.

Reviewer #1: Yes

Reviewer #2: Yes

**********

3. Has the statistical analysis been performed appropriately and rigorously?

Reviewer #1: Yes

Reviewer #2: Yes

**********

4. Have the authors made all data underlying the findings in their manuscript fully available?

The PLOS Data policy requires authors to make all data underlying the findings described in their manuscript fully available without restriction, with rare exception (please refer to the Data Availability Statement in the manuscript PDF file). The data should be provided as part of the manuscript or its supporting information, or deposited to a public repository. For example, in addition to summary statistics, the data points behind means, medians and variance measures should be available. If there are restrictions on publicly sharing data—e.g. participant privacy or use of data from a third party—those must be specified.

Reviewer #1: Yes

Reviewer #2: Yes

**********

5. Is the manuscript presented in an intelligible fashion and written in standard English?

PLOS ONE does not copyedit accepted manuscripts, so the language in submitted articles must be clear, correct, and unambiguous. Any typographical or grammatical errors should be corrected at revision, so please note any specific errors here.

Reviewer #1: Yes

Reviewer #2: Yes

**********

6. Review Comments to the Author

Please use the space provided to explain your answers to the questions above. You may also include additional comments for the author, including concerns about dual publication, research ethics, or publication ethics. (Please upload your review as an attachment if it exceeds 20,000 characters)

Reviewer #1: The minor issues in the manuscript have been corrected except stating in the abstract that this is a developing country.

Reviewer #2: (No Response)

**********

7. PLOS authors have the option to publish the peer review history of their article (what does this mean?). If published, this will include your full peer review and any attached files.

If you choose “no”, your identity will remain anonymous but your review may still be made public.

Do you want your identity to be public for this peer review? For information about this choice, including consent withdrawal, please see our Privacy Policy.

Reviewer #1: No

Reviewer #2: No

**********

4 Aug 2022

PONE-D-22-03245R2

Osteoporotic hip fracture – comorbidities and factors associated with in-hospital mortality in the elderly: a nine-year cohort study in Brazil

Dear Dr. Peterle:

I'm pleased to inform you that your manuscript has been deemed suitable for publication in PLOS ONE. Congratulations! Your manuscript is now with our production department.

If your institution or institutions have a press office, please let them know about your upcoming paper now to help maximize its impact. If they'll be preparing press materials, please inform our press team within the next 48 hours. Your manuscript will remain under strict press embargo until 2 pm Eastern Time on the date of publication. For more information please contact onepress@plos.org.

If we can help with anything else, please email us at plosone@plos.org.

Thank you for submitting your work to PLOS ONE and supporting open access.

Kind regards,

PLOS ONE Editorial Office Staff

on behalf of

Professor Robert Daniel Blank

Academic Editor

PLOS ONE