Malnutrition is Common and Increases the Risk of Adverse Medical Events in Older Adults With Femoral Fragility Fractures

Abstract

Background

Femoral fragility fractures are one of the most common injuries managed by orthopedic surgeons. Malnutrition influences the poor outcomes observed in this population. Our purpose was to assess the annual trends of malnutrition diagnosis and determine risk factors for malnutrition and complications in patients 65 years and older presenting with femoral fragility fractures. We hypothesized that malnutrition would increase the risk of postoperative wound infection, wound dehiscence, non-union, and mortality.

Methods

The PearlDiver database was reviewed from 2010 to 2020. Patients ≥ 65-years-old with femur fractures treated with operative fixation were identified by CPT code. A preoperative diagnosis of malnourished state was defined by ICD-9 and ICD-10 codes and patients were divided into malnourished and non-malnourished cohorts. Patients were tracked for one year following operative fixation of a femoral fragility fracture for the occurrence of infection, wound dehiscence, nonunion and mortality. The rates of these complications were compared between malnourished and nonmalnourished cohorts.

Results

There were 178,283 total femoral fragility fractures identified in patients aged 65-years or older. The overall prevalence of malnutrition diagnosis in this geriatric population was 12.8%. Documented malnutrition in femoral fragility fractures increased from 1.6% to 32.9% from 2010-2020 (P<0.0001). Compared to patients without malnutrition, patients with malnutrition are at increased risk of mortality (OR 1.31, 95% CI 1.2558 – 1.3752, p < 0.0001), are more likely to develop a wound infection (OR 1.49; 95% CI 1.252 – 1.7626; p < 0.0001), more likely to have a wound dehiscence (OR 1.55; 95% CI 1.3416 – 1.7949; p < 0.0001), and more likely to develop non-union (1.89; 95% CI 1.6946 – 2.1095; p < 0.0001). Multiple demographic variables were associated with malnutrition diagnosis including higher age, higher Charlson Comorbidity Index, female sex, dementia, and institutionalization. Parkinson’s disease, feeding difficulty and institutionalization demographic variables had the highest risk of malnutrition.

Conclusion

The current study found that malnutrition diagnosis significantly increases the risk of adverse medical events in elderly adults with femoral fragility fractures. The rates of malnutrition increased steadily from 2010-2020. This trend is likely a result of increased awareness and testing for malnutrition, not reflecting an actual increased prevalence of malnutrition. Multiple expected demographic variables are associated with diagnosis of malnutrition.

Level of Evidence: III

Introduction

Musculoskeletal trauma in the elderly population is common with an estimated 250,000 patients presenting annually with hip fractures.¹ The prevalence of malnutrition ranges from 24% to 88% depending on the population studied and how malnutrition is defined.^2,3 Malnutrition is a potentially modifiable risk factor leading to increased rates of infection, delayed wound healing, hospital length of stay and mortality.^3-5

Large database studies have been used to document the impact of malnutrition diagnosis on adverse medical events after orthopedic surgery. A study of 49,603 patients who underwent primary total hip and knee arthroplasty identified in the American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database found a malnutrition rate of 4% defined as an albumin concentration < 3.5 g/ dL and reported increased relative risks of surgical site infection, pneumonia, and renal insufficiency.⁶ Similarly, a systematic review and meta-analysis of the association of malnutrition and infection (periprosthetic and surgical site) in > 250,000 patients after total joint arthroplasty reported increased risk of both prosthetic joint infections and surgical site infections among malnourished patients.⁵ Using the Nutritional Risk Screening tool to prospectively assess 1055 orthopedic and trauma patients 22% were found to be at risk for malnutrition.⁴ Patients at risk for malnutrition also had higher rates of adverse events and tended toward higher rates of local and systemic infections and decreased wound healing. To our knowledge malnutrition in the elderly population with femoral fragility fracture has not been investigated as a risk factor for adverse medical events in the PearlDiver Database.

Further investigation is needed to determine how the diagnosis of malnutrition impacts clinically important outcomes in the elderly femoral fragility fracture population. The goals of this study are to (1) Determine the prevalence of malnutrition in elderly patients with femoral fragility fractures, (2) Identify risk factors for malnutrition, and (3) Assess the impact of malnutrition diagnosis on postoperative complications in this population.

Methods

All data was collected in the PearlDiver Database. This database is a vault of millions of Health Insurance Portability and Accountability Act compliant patient medical records with the ability to capture diagnostic, procedural and prescription data. Researchable coding includes International Classification of Diseases Ninth and Tenth Revisions (ICD-9 and ICD-10) for both diagnosis and procedure as well as Current Procedural Terminology (CPT) and National Drug Code (NDC).

The PearlDiver dataset was retrospectively queried from 2010 to the first quarter of 2020 to determine the incidence of femoral fragility fractures. Operative femur fractures were identified by CPT codes (OTA 31, 32, and 33).⁷ Patients were included with age 65-years-old or older with proximal, midshaft or distal femur fractures. Table 1 lists the CPT codes queried. Malnutrition diagnosis and postoperative variables including wound infection, wound dehiscence and non-union were identified by ICD-9 and ICD-10 codes. Patients were then separated into two cohorts, malnourished and nonmalnourished, and were evaluated for one year following operative fixation of their femoral fragility fracture for the development of complications. Mortality rate is not a directly identifiable variable in the database. Survival was indirectly determined by their continued enrollment in the database at least 30 days after surgery.

Table 1.

Operative Femur Fracture Identified by CPT Code

CPT code	Description
27235	Percutaneous skeletal fixation of femoral fracture, proximal end, neck, in situ pinning of undisplaced or impacted fracture
27236	Open treatment of femoral fracture, proximal end, neck, internal fixation or prosthetic replacement
27244	Treatment of intertrochanteric, peritrochanteric, or subtrochanteric femoral fracture; with plate/screw type implant, with or without cerclage
27245	With intramedullary implant, with or without interlocking screws and/or cerclage
27506	Open treatment of femoral shaft fracture, with or without external fixation, with insertion of intramedullary implant, with or without cerclage and/or locking screws
27507	Open treatment of femoral shaft fracture with plate/ screws, with or without cerclage
27511	Open treatment of femoral supracondylar or transcondylar fracture without intercondylar extension, includes internal fixation, when performed
27513	Open treatment of femoral supracondylar or transcondylar fracture with intercondylar extension, includes internal fixation, when performed
27514	Open treatment of closed or open femoral fracture, distal end, medial or lateral condyle, includes internal fixation, when performed

Statistical Analysis

The frequencies (percentages) of malnutrition and complications including infection, wound dehiscence, nonunion and mortality were determined. Trend analysis to determine the prevalence of malnutrition over time was done utilizing Poisson regression. The relationships between the odds of complications and malnutrition were determined using logistic regression with and without adjustment for Charlson Comorbidity Index (CCI scores). We also explored the relationships between participant factors and odds of malnutrition with univariate logistic regression.

Results

There were 178,283 total femur fractures identified. The overall prevalence of malnutrition in this geriatric population was 12.8% between 2010 and the first quarter of 2020. During this same period of time the prevalence of malnutrition diagnosis significantly increased from 1.6% to 32.9% (p <0.001). (Figure 1). Demographic data was compared for the malnutrition and non-malnutrition cohorts. The malnourished cohort had a mean age of 78-years-old compared to the non-malnourished cohort which had a mean age of 75-years-old (p < 0.001) and the malnourished cohort was 73% female compared to 72% in the non-malnourished cohort among other significant variables (p = 0.0014).

Figure 1.

Malnutrition Prevalence for Patients with Femur Fractures

Odds ratios (OR) for adverse medical events were determined. Compared to non-malnourished patients, malnourished patients have an increased risk of mortality (non-continuous enrollment) (OR 1.31; 95% CI 1.2558 – 1.3752, p < 0.0001), are more likely to develop a wound infection (OR 1.49; 95% CI 1.252 – 1.7626; p < 0.0001), are more likely to have a wound dehiscence (OR 1.55; 95% CI 1.3416 – 1.7949; p < 0.0001), and are more likely to have a non-union (OR 1.89; 95% CI 1.6946 – 2.1095; p < 0.0001) (Table 2).

Table 2.

Odds Ratios for Mortality and Adverse Medical Events

	Univariate			Multivariate
	OR	95% CI	P-value	OR	95% CI	P-value
Mortality	1.55	1.4798- 1.6145	<0.0001	1.31	1.2558- 1.3752	<0.0001
Wound Infection	1.59	1.3478- 1.8825	<0.0001	1.49	1.252- 1.7626	<0.0001
Wound Dehiscence	2.20	1.1133- 4.3375	0.0343	1.55	1.3416- 1.7949	<0.0001
Non-union	1.99	1.7886- 2.2119	<0.0001	1.89	1.6946- 2.1095	<0.0001

Potential risk factors for malnutrition were also evaluated including: age, sex, CCI, diabetes, obesity, dementia, depression, Parkinson’s disease, constipation, dysphagia, feeding difficulty, cognitive decline, institutionalization, tobacco use and alcohol use. Three factors were associated with a greater than two times risk for malnutrition and include Parkinson’s disease (OR 2.19; 95% CI 2.0486 – 2.3435; p < 0.001), feeding difficulty (OR 2.09; 95% CI 1.9192 – 2.0854; p < 0.0001) and institutionalization (OR 2.33; 95% CI 2.1678 – 2.5038; p < 0.0001) (Table 3).

Table 3.

Odds Ratios for Risk Factors of Malnutrition

	Malnutrition*	No Malnutrition*	Odds Ratio	95% CI	P-value
Age	78 (74-80)§	75 (73-78)§			<0.0001
CCI	4.34†	2.56†			<0.0001
Female	16636	111380	1.0521	(1.0199-1.0853)	0.0014
Diabetes	11704	63836	1.5014	(1.4602-1.5438)	<0.0001
Obesity	4274	18264	1.7243	(1.6622-1.7887)	<0.0001
Dementia	7160	25651	1.7197	(1.6186- 1.8268)	<0.0001
Tobacco	6059	26128	1.7816	(1.7251-1.84)	<0.0001
Alcohol	1792	6414	1.9734	(1.869-2.0835)	<0.0001
Depression	12202	60241	1.8042	(1.7546- 1.8553)	<0.0001
Parkinson’s Disease	1168	3723	2.1911	(2.0486-2.3435)	<0.0001
Constipation	10406	50477	1.7333	(1.6853-1.7827)	<0.0001
Dysphagia	7938	37328	1.6798	(1.6308-1.7303)	<0.0001
Cognitive decline	3124	11418	1.9935	(1.9110-2.0796)	<0.0001
Institutionalization	1027	3072	2.3298	(2.1678-2.5038)	<0.0001
Feeding difficulty	746	2471	2.0854	(1.9192- 2.0854)	<0.0001

*Values given are the number of patients. §Represents patient age. †Charlson Comorbidity Index.

Discussion

This study was designed to determine the prevalence of malnutrition in elderly patients treated with operative fixation for femoral fragility fractures. We report an overall 12.8% rate of malnutrition with significant increase in diagnosis from 1.61% to 32.90% from 2010 to the first quarter of 2020. This is not an unexpected finding given the increased awareness of malnutrition diagnosis, readily available screening tools,^3,8 and serum chemistries. However, this apparent increasing prevalence is likely related to the increased awareness of malnutrition and not an actual increase in prevalence in this patient population. The actual prevalence is likely closer to the results we found in 2020 as awareness and documentation increased. Rates of complications including mortality (non-continuous enrollment), wound infection, wound dehiscence and non-union were all significantly increased in malnourished patients treated operatively for femoral fragility fractures. Many risk factors for malnutrition were identified, including Parkinson’s disease, feeding difficulty and institutionalization as the greatest demographic risk factors for malnutrition.

The rates of malnutrition in the literature among patients undergoing musculoskeletal surgical procedures have a wide range. In 1982 Jensen et al. studied 129 patients undergoing orthopedic procedures and reported an average rate of 42% using anthropometric measurements (triceps skinfold, arm circumference) and biochemical testing (serum albumin <3.5 g/dl) as the definition for malnutrition. Subgroup analysis revealed a malnutrition rate of 28% among patients undergoing total hip arthroplasty and 59% among poly traumatized patients.⁹ A study of the ACS-NSQIP database from 2005-2013 identified 4,655 patients that underwent total shoulder arthroplasty finding a rate of 7.6%.¹⁰ A study of 327 patients from an urban U.S. level 1 trauma center with isolated hip fractures found the prevalence rate of malnutrition to be 17.5% when defined by an albumin level <3.5 g/dL.¹¹ Similar to these studies we also report prevalence in our population. Unlike other studies, however, we also provide a trend of malnutrition over a 10-year period of time and report a significant increasefrom 1.6%-32.9%. The prevalence found in later years likely better reflects this population as awareness of the diagnosis has increased.

Increased complication and mortality rates have been previously described in the literature. In 2017 Chung et al. retrospectively studied 12,373 patients with hip fractures and found a 2-fold increased risk of postoperative complications and mortality.¹² More recently in a study of 20,278 patients with hip fractures from the ACS-NSQIP database reported increased odds ratios for deep wound infection, any infection, reintubation, prolonged length of stay, readmission, reoperation, in-hospital mortality and 30-day mortality among malnourished patients. In this study malnutrition diagnosis was determined with an albumin level <3.5 g/dl.¹³ We report similar findings and reinforce the results of prior studies demonstrating increased rates of mortality and post-operative complications in malnourished surgical patients.

Many known risk factors exist for malnutrition in elderly patients. A systematic review of the literature based on longitudinal data from MEDLINE between 2000 and March 2015 found 15 significant risk factors for malnutrition in the older population (aged ^≥ 65-years-old).¹⁴ We confirm many of the same risk factors in elderly adults with femoral fragility fractures including: age, sex, CCI, diabetes, obesity, dementia, depression, Parkinson’s disease, constipation, dysphagia, feeding difficulty, cognitive decline, institutionalization, tobacco use and alcohol use. With a predicted global increase in life expectancy between 2010 and 2050¹⁵ we also expect an increase of older adults with malnutrition.

Increased mortality rates after hip fractures have been well studied and previously demonstrated.^16,17 Diaphyseal and distal femur fractures have not been as extensively studied but a recent study of 11,799 femoral fractures found these fractures may have similar early mortality risks as hip fractures.¹⁸ We included all femoral fragility fractures as these patients have a similar prevalence of adverse medical event.

Our study adds to a growing body of literature about the significant impact malnutrition plays in surgical patients and specifically among elderly patients with femoral fragility fractures. It is important to recognize that malnutrition is a potentially modifiable risk factor and provides us an opportunity to improve the outcomes of this patient population.

Limitations

Limitations of this study should be acknowledged. First, the PearlDiver database contains over 4 billion patient records obtained from analysis of private insurance claims from Humana and United Healthcare, as well as government claims from Medicare.¹⁹ As a private analytics database it is not a random sample so data results and generalizations must be cautiously interpreted. Second, this is a retrospective study and the results are dependent on the completeness and accuracy of the data collected by PearlDiver. Furthermore, the results are dependent on diagnosis coding accuracy. Medical coding is very complex and the transition from ICD-9 to ICD-10 on October 1, 2015 only increased that complexity. For example, ICD-9 includes one code for fracture non-union, however, ICD-10 includes over 200 codes for fracture non-union of the femur. Third, mortality cannot be directly measured in the PearlDiver database so continuing with insurance for 30 days post-operatively served as a surrogate. It is likely that patients discontinued their insurance within 30 days post-operatively for reasons other than death. Other studies have used these criteria for mortality in the PearlDiver database.^20,21

Conclusion

We found that malnutrition significantly increases the risk of adverse medical events in the elderly population with femoral fragility fractures. The rates of malnutrition increased steadily from 2010-2020, inferring an increased risk of malnutrition in 2020 compared to 2010, however, this trend is likely a result of increased awareness and testing for malnutrition and may not actually reflect increased rates of malnutrition. Diagnosis of malnutrition is associated with mortality (non-continuous enrollment), wound complications, and non-union. Significant risk factors for malnutrition in this study are Parkinson’s disease, feeding difficulty and institutionalization among others.