Overview

Hip fractures rank in the top ten of all impairments worldwide in terms of loss in disability-adjusted years for people 50+ years old.¹ Hip fracture rates in elderly population in the United States have decreased from 901 per 100,000 population in 1993 to 776 per 100,000 population in 2003; however, the absolute number of hip fracture hospitalizations increased 19 percent, from 261,000 to 309,500 during the same time period.² This number is expected to continue to rise due to increased life spans and an aging baby boom generation; by 2030, 71 million older adults will account for roughly 20 percent of the U.S. population.⁵³ The lifetime incidence of hip fracture is 17 to 22 percent for 50 year old women, 6 to 11 percent for men, and, while the rate for women stabilizes, there is an age-specific increase in incidence rates for men.^54,55

Ninety percent of hip fractures result from a simple fall.³ Although it might be argued that hip fracture incidence may change in the future due to recent drug interventions for osteoporosis,⁴ efforts to prevent hip fractures are unlikely to have a substantial impact on hip fracture incidence.^4–7 In fact, many studies have found considerable overlap in bone mineral density and bone mass between hip fracture patients and their controls, suggesting that other factors are strong contributors to hip fracture risk. These factors include body size characteristics, decreased muscular strength, inactivity, impaired cognition, impaired perception and vision, environmental circumstances, chronic illnesses and drugs that may contribute to the propensity to fall, such as psychotropic anxiolytic/hypnotic drugs, barbiturates, opioid analgesics, antihypertensives, anticonvulsants, caffeine, tranquilizers, sedatives, antidepressants.^4,56

Consequences of hip fractures are significant in terms of lives lost and the associated negative impacts on hip fracture patients’ functioning and quality of life.⁸ One year mortality for patients after a hip fracture is approximately 20 percent, with men, patients older than 75 years, and nursing home residents at higher risk,^4,9–11 as much as a three-fold increase in the first year after fracture.⁵⁷ However, it is not clear how much of that mortality is due to the fracture and its sequellae and how much to the underlying frailty that may predispose a person to a fracture.¹² Morbidity associated with hip fractures may include serious complications, such as deep vein thrombosis, muscular deconditioning, postoperative infection, pain, and loss of mobility.⁴ Among patients who were living independently prior to a hip fracture, only about half are able to walk unaided after fracture,^13,14 and about one-fifth require long-term placement in a care facility.^15,16

The goal of treatment for hip fractures is to return patients to their pre-fracture level of function.^6,20,21 Treatment options for hip fracture patients depend on the location and pattern of the fracture, patient characteristics, and available Food and Drug Administration (FDA) approved devices. There is a growing body of literature on treatment options and intermediary and patient post-surgical treatment outcomes, including several systematic reviews (see Appendix A). However, no comprehensive organization of the evidence as it specifically relates to patient outcomes currently exists. The current body of systematic literature reviews generally relies on intermediate outcomes such as complications, reoperation rates, and device failure in generating hip fracture treatment recommendations. The aim of the present project was to conduct a systematic review and synthesize the evidence of the effects of surgical treatments of fractures of the hip on patient post-treatment outcomes, which to date have received only modest attention, but which are central to the patient’s experience.

Types of Hip Fractures

Hip fractures occur in the proximal (upper) portion of the femur, just outside the area where the femoral head (ball) meets the acetabulum (socket) within the pelvis. (Figure 1) Hip fractures are generally classified into three major types, depending on the specific location of the fracture: femoral neck, intertrochanteric, and subtrochanteric fractures. The term pertrochanteric hip fracture may also be used in hip fracture literature and refers to a more inclusive set of extracapsular fractures, including intertrochanteric, subtrochanteric, and mixed fracture patterns.⁶ The term hip fracture describes fractures of the upper thigh or femur; fractures of the acetabulum (socket) and other pelvic fractures are not considered to be hip fractures and will not be discussed in this report.

Figure 1

Diagram of right hip as viewed from the front. Fractures of the upper portion of the femur, are hip fractures. Hip fractures do not involve the acetabulum (socket). From Zuckerman, 1996. Used with permission.

Femoral neck fractures occur in the narrowed section of the upper femur that lies between the rounded femoral head and bony projections called trochanters. Most femoral neck fractures occur within the capsule that surrounds the hip joint and are therefore termed intracapsular. The blood supply to the femoral head is entirely dependent upon a series of arteries that pass through the femoral neck region. Therefore, fractures of the femoral neck can entirely disrupt the blood supply to the femoral head, resulting in increased rates of major healing complications such as fracture nonunion, or death of the femoral head bone tissue, called osteonecrosis, or avascular necrosis. Femoral neck fractures are further grouped into nondisplaced and displaced fractures by the alignment of the fractured segments in relation to the original anatomic position of the femur.

Intertrochanteric fractures occur more laterally than femoral neck fractures, in the area between the greater and lesser trochanters. The trochanters are bony projections where major hip muscles attach. Intertrochanteric hip fractures occur outside of the joint capsule and are therefore extracapsular. Unlike the complicated blood supply through the femoral neck, intertrochanteric hip fractures involve bone with good local blood supply. However, these fractures are complicated by the pull of the hip muscles on the bony muscle attachments, which can exert competing forces against fractured bone segments and pull them out of alignment. Thus, the healing complications for intertrochanteric fractures are often different from those of femoral neck fractures, and are more likely to include shortening of the length of the femur or healing of the fracture in a misaligned position (malunion). Intertrochanteric fractures may be further grouped into stable and unstable fractures, depending on the location, number, and size of the fractured bony segments.

Subtrochanteric fractures occur at or below (distal to) the level of the lesser trochanter in the proximal portion of the femur. Isolated subtrochanteric fractures occur in the area between the upper border of the lesser trochanteric to 5 cm below it, toward the knee. Subtrochanteric fractures may include only a short linear section of the proximal femur or may be part of a larger fracture pattern that involves both the intertrochanteric and subtrochanteric sections of the femur. The blood supply to the bone of the subtrochanteric region is not as good as the blood supply to the bone of the intertrochanteric region and thus heals more slowly. Subtrochanteric fractures are also subject to competing forces exerted by muscular attachments on the femur that tend to pull the fractured fragments out of alignment.

Orthopaedic surgeons differ on their definition of subtrochanteric fractures, and may also consider fractures that extend further toward the knee to be subtrochanteric. There are at least 15 classification systems in the literature for subtrochanteric fractures, the Seinsheimer system being the most frequently used.⁵⁸

Michelson et al.,¹⁷ found that the distribution of the types of hip fractures within the U.S. population is 49 percent intertrochanteric, 37 percent femoral neck, and 14 percent subtrochanteric, and these estimates are relatively consistent across authors.^11,18,19 The proportions may change over time as demographics change,⁴ and proportions are also reliant on the classification system used and the preferences of surgeons in labeling fracture patterns. This is a problem particularly for subtrochanteric fractures, where estimates of the percent of hip fractures that are subtrochanteric vary widely, depending on the classification system used.⁵⁸ Within the elderly population, the incidence of subtrochanteric fractures is estimated to be about 10 percent of all hip fractures.^19,59

Table 1 summarizes the major differences in femoral neck and trochanteric fractures that are important to the management of these fractures. Nonoperative management of a hip fracture may be indicated for nonambulatory, institutionalized patients, patients with serious medical conditions, and where surgical intervention is considered too risky and nursing care can be accomplished without causing significant pain to the patient. However, the vast majority of patients with a hip fracture are treated surgically.

Table 1

Hip fracture management characteristics.

Surgical Treatment of Hip Fractures

The short term goal of surgical treatment is to stabilize the hip fracture enough to withstand early mobilization and weight bearing, which prevents complications due to prolonged bed rest and aids in fracture healing. This type of surgery is generally based on the fracture pattern and patient characteristics.

Fracture healing is best accomplished when the edges of the fractured bone are perfectly realigned into a normal anatomic position (fracture reduction), and the fractured edges contact and press firmly upon each other (impaction) in good alignment, which stimulates bone healing. This is particularly important in elderly patients who are more likely to have low bone density and other medical issues that impair their ability to heal a fracture and recover hip function. Many hip fractures are treated with internal fixation, which is a general class of metal implants placed in or anchored directly alongside of bone to hold the fractured bony segments in alignment until the fracture can fully heal. Internal fixation devices are most often left in place after hip fractures heal, but they may be removed if painful. Severe complications in fracture healing are more likely to occur if the fracture was not sufficiently reduced before an internal fixation implant was placed, or if the implant was not placed and anchored properly.²⁰ If the optimal realignment of a fracture is not possible or practical for other reasons, then a partial or total hip replacement procedure (arthroplasty) is considered. An arthroplasty by its very nature replaces the function of the affected joint and is hence devoid of fracture healing complications. For arthroplasty, implant placement and surgical approach (where the incision is made) are important in predicting which complications are most likely to occur.²⁰

Femoral Neck Fractures

Nondisplaced fractures. Fractures of the femoral neck that are found on initial x-rays to be in normal or near-normal alignment are called nondisplaced femoral neck fractures by most surgeons (Figure 2). About one-third of femoral neck fractures are nondisplaced.^60,61 The treatment of nondisplaced femoral neck fractures is fairly uncontroversial and generally involves internal fixation.⁶² Multiple stabilizing screws are placed from the outside (lateral) portion of the proximal femur through the fractured femoral neck, and anchored into the bone of the femoral head. With the fracture aligned and stabilized with screws, these fractures commonly heal without complications.

Figure 2

Garden classification for femoral neck fractures. Type I and II are considered nondisplaced fractures. Type III and IV are considered displaced. Used with permission.

Displaced fractures. Fractures of the femoral neck that are found to be moderately to severely misaligned on initial x-rays are called displaced femoral neck fractures. More controversy surrounds the treatment for displaced femoral neck fractures.^63,64 Elderly patients with displaced femoral neck fractures may be treated with internal fixation or arthroplasty, either hemiarthroplasty, or total hip arthroplasty.^62,64 Device selection depends on multiple factors, such as the patient’s general health status, age, functional abilities, cognitive status, and surgeon preference.^6,20,62 When selecting the type of implant that is best for a patient, surgeons must simultaneously weigh the likelihood of achieving a satisfactory fracture reduction, the risk of nonunion of the fracture, the patient’s medical condition and activity level, and the possibility of osteonecrosis of the femoral head if the blood supply through the femoral neck is damaged. Maintaining a patient’s own femoral head through a satisfactory reduction is a treatment goal whenever possible.⁶² Delays between sustaining a fracture and surgery may reduce the chance of uncomplicated healing of the femoral neck, and thus is also a consideration in treatment decisions.

Internal fixation. Internal fixation is often used in situations where the patient is relatively young, active, has a fracture pattern that is likely to be able to maintain reduction of the fracture, and has sufficient bone quality to anchor and hold the implant in place without allowing the fracture to re-displace during healing.⁶² In addition to the screws mentioned for nondisplaced fractures, plate and screw combinations may be used as well, to hold the reduced fracture in position. However, no internal fixation device can compensate for an inadequately reduced fracture.²⁰

Arthroplasty. Treatment choices shift to arthroplasty for displaced femoral neck fractures in the physiologically older patient with poor bone quality. Two types of arthroplasty are available: hemiarthroplasty and total hip arthroplasty (THA). Hemiarthroplasty is the replacement of the femoral head segment of the upper femur with an artificial implant. The patient’s own acetabulum is not replaced, hence the term hemi-, or half of a new hip. THA is the prosthetic replacement of the entire hip joint, both the femoral head and the acetabulum within the pelvis. Regardless of the type of hemiarthroplasty, hip pain as a result of acetabular erosion is likely to develop in due course of time (2–5 years or sooner). With this in mind, the patient’s physiological status and likely life span are important factors in decisionmaking. Therefore, hemiarthroplasty is considered better suited to less active elderly patients who are expected to place low activity demands on their prosthesis, and whose acetabulum does not have extensive degenerative or arthritic type changes that could be aggravated by this type of implant. THA is used for patients who are more active at the time of their hip fracture, or whose hip joint already had severe degenerative changes.

Two types of hemiarthroplasty implants are available: unipolar or bipolar. Unipolar hemiarthroplasty consists of an implant with an artificial femoral head that acts as the ball of the hip joint, and a metallic rod, or stem, that is inserted into the innermost portion of the femur (intramedullary canal) to hold the entire device firmly in place. Some patients with early unipolar hemiarthroplasty designs experienced hip pain due to erosion of the acetabulum, or thigh pain due to loosening of the stem in the femoral canal. Later designs included cement to anchor the prosthetic stem into the mid-femur. Another design modification is the use of a bipolar hemiarthroplasty. The bipolar design incorporates a smaller femoral head which fits inside a second larger ball or shell, which together act as the artificial femoral head. Bipolar designs attempt to cut down on the amount of wear and tear the prosthesis causes to the patient’s normal hip joint by allowing the smaller inside head to accomplish more movement than the outer rounded portion that contacts the patient’s acetabulum. Bipolar hemiarthroplasty designs cost about $1,000 more per implant than unipolar implants. Bipolar prostheses can also be converted to total hip replacements if required.

In THA, the artificial femoral head is attached to an anchoring metallic rod that extends down within the femur to the mid thigh area, similar to hemiarthroplasty. The prosthetic femoral head moves within the artificial socket to allow for near normal hip movements after healing. Various designs exist and design improvements are aimed at minimizing the amount of fine, particulate erosion that can occur within the artificial joint that can diffuse into the surrounding tissue and cause irritation or other undesirable changes. The same issues that can arise with the femoral stem component for hemiarthroplasty can also occur for total hip arthroplasty, since the stem component options are the same.

Current issues. Within these broad bounds, considerable controversy regarding which forms of surgical treatments are best for which patients remains.^49,62,63 Internal fixation saves the original femoral head but is more likely to require operative revision.^28,65,66 THA has a lower revision rate but is subject to dislocation, or slippage of the artificial femoral head out of the artificial socket. Dislocation is less common with hemiarthroplasty, but revision rates for hemiarthroplasty are somewhere between internal fixation and THA revision rates. However, technical failures and revision rates are affected by more than device design. A recent study that tested a physiologic status score as a decision support for whether internal fixation or hemiarthroplasty was used as surgical treatment found that neither age nor physiologic status was predictive of internal fixation failure.⁴⁹ However, when independent reviewers assessed the surgical treatments for technical failure, defined as inadequate reduction or improperly placed implants, technical failure was reported in 14 percent of internal fixations cases, in only 2 percent of hemiarthroplasties, and these errors in surgical technique correlated with worse clinical outcomes.⁴⁹

Pertrochanteric Fractures

Pertrochanteric fractures, which include intertrochanteric and subtrochanteric fractures, are typically treated with internal fixation. Since pertrochanteric fractures do not disrupt the blood supply to the femoral head, internal fixation is the treatment of choice, unless the entire hip joint is severely deformed due to advanced degenerative changes, or less commonly, when the fracture extends proximally into the femoral neck area.

Types of internal fixation. The two major forms of internal fixation for pertrochanteric fractures are extramedullary and intramedullary implants.

Extramedullary. Extramedullary devices attach externally along the outside of the upper femur. One of the earliest and most commonly used extramedullary devices is the sliding hip screw.⁶⁷ The term is somewhat misleading, since this category of devices actually has two main parts: a metal plate and multiple screws. The plate of a sliding hip screw is affixed along the outside of the femur with multiple screws that cross the upper femur to hold it in place. The upper portion of the device consists of a separate longer and thicker screw called a lag screw that is fit lengthwise from the outside, upper femur through the femoral neck into the hard bone of the femoral head. The term sliding hip screw describes the unique, intended action that allows the surgically-aligned fracture segments to impact firmly on each other to aid fracture healing, particularly when a patient stands or walks.

There are a number of versions of sliding screw/plate devices. Multiple modifications in designs have been engineered to address different forms of fixation failures and design flaws, such as the amount of sliding allowed where the lag screw meets the plate, the direction of sliding allowed, the length, design, and best placement of the lag screw, the number of nonlag screws needed and where to best position them, and variations in plate length and the number of anchoring screws needed. Fixation forms may vary in terms of rigidity of the device and the types and locations of the specific fixation to the femur in order to maintain optimal fracture reduction and necessary fracture impaction without the device failing or pulling out of the bone.

Intramedullary. Intramedullary internal fixation devices use the same type of anchoring lag screw that is used in the sliding screw/plate devices, and the controlled bony impaction at the fracture site is accomplished with the same dynamic motion as in the sliding hip screws. However, unique to this class of implants is the portion that controls the fracture, which is placed inside the canal of the femur, rather than alongside it, hence the term intramedullary. The devices are commonly called intramedullary nails (IMN), since the femoral portion of the device is essentially a very large surgical nail. The procedure requires a smaller incision than plate/screw procedures, near the top of the hip. Once the fracture is properly aligned by the surgeon, the IMN is slid into the central portion of the femur from the hip toward the knee, typically ending in the middle to lower third of the femur, but always above the knee.

The theoretical biomechanical advantage of IMN is that the central placement of the device within the femur leaves a shorter distance between the device and the weight bearing axis of the femur, thereby attempting to lessen the forces that pull fracture fragments out of alignment until the hip fracture heals. Because of the smaller incision, surgical procedures with IMN tend to have less blood loss. IMN devices have their own set of complications, including unintentional fractures intraoperatively as the surgeon places the nail, and later fracture of the femur at the distal end of the nail. As with plate/screws devices, multiple generations of IMN devices have been engineered, varying nail length, width, and curvature, among others, in attempts to improve treatment outcomes.

Stable and unstable fractures. Surgeons use multiple classification systems to identify intertrochanteric hip fracture patterns. The most commonly used are the AO/OTA (Arbeitsgemeinschaft fur Osteosynthesefragen (Association for the Study of Internal Fixation)/Orthopedic Trauma Association) classification (Figure 3) and the Evans classification, as modified by Jensen (Figure 4). The AO/OTA classification has nine distinct intertrochanteric subtypes, whereas the Evans-Jensen classification has five. Unlike bone in the mid portion of the femur, the bony anatomy of the intertrochanteric area is complex and fracture patterns vary widely. For randomized clinical trials, fracture classifications are used to group several fracture subtypes together that have similar treatment and healing issues in order to make general conclusions about which fracture patterns do better with which category of device. Typically, intertrochanteric fracture patterns are aggregated into stable and unstable groupings, and these fractures behave quite differently in terms of how they heal with an optimally placed device.

Figure 3

AO/OTA classification of proximal femur fracture. AO/OTA classification of proximal femur fractures. The A1 fracture would routinely be considered a simple, stable intertrochanteric femur fracture. The A2 fracture is characterized by a similar fracture (more...)

Figure 4

Evans/Jensen classification for stable/unstable intertrochanteric fractures. Used with permission.

Isolated subtrochanteric fractures, or intertrochanteric fractures with subtrochanteric extension, are generally considered to be unstable fractures. Subtrochanteric fractures are complicated by the multiple strong muscular forces acting across the site of the fracture. Biomechanical studies have found the region one to three inches below the lesser trochanter, a small bony knob on the upper and inner portion of the femur, to be the most highly stressed region of the body,⁶⁸ which increases the possibility of implant failure and malunion of the fracture. Device options include similar internal fixation devices that are used for isolated intertrochanteric fractures. Intramedullary nails may be indicated for certain subtypes of fractures that are particularly unstable.⁶⁹ Sliding hip screw devices are not well suited for subtrochanteric and reverse oblique intertrochanteric fractures as they may allow unacceptable displacement of the femur shaft. Again, as for all surgical decisions, surgeon experience and familiarity with the devices must also be considered.^68,70

Current issues. The choice between extramedullary or intramedullary devices for pertrochanteric fracture treatment among surgeons remains controversial,^21,32,67,71 and significant variation in surgeons’ device choice for pertrochanteric fractures exists.^19,67 For stable fracture patterns, the literature suggests that the sliding plate/screw devices remain the treatment of choice,^21,30,72 although surgeon practice varies widely.^19,67 However, for unstable fractures, consensus has not been reached, with the possible exception of the reverse oblique fracture for which the intramedullary nail is better suited.^21,32,69 Intramedullary devices cost approximately $1,000 more per implant than extramedullary devices.^19,21,67

Additionally, there is no consensus among orthopedic surgeons internationally as to which subtypes of the current AO/OTA or Evans-Jensen classification systems are considered to be unstable fractures, and may therefore warrant special device considerations. This is not an issue with femoral neck fracture trials, since surgeons agree on the aggregation of femoral neck fracture subtypes into displaced or nondisplaced categories. For pertrochanteric fractures, this lack of consensus causes considerable challenges in the interpretation of implant-related outcomes within the existing literature, since investigators select different fracture pattern subtypes as unstable.

As with femoral neck fractures, surgeon skill has been found to be a strong predictor of fixation failure.⁷³ Accurate placement of the lag screw is one of the strongest predictors of whether or not the lag screw will pull out of the femoral head (cut-out) and result in implant failure.^74,75

Key Questions

This review was asked to address the following key questions:

• Key Question 1 – What is the relationship between patient variables (e.g., demographic factors, comorbidities), the type of fracture (i.e., intertrochanteric, subtrochanteric, subcapital) and post-treatment outcomes (e.g., pain, mobility, mortality)?
• Key Question 2 – What is the relationship between the type of fracture (i.e., intertrochanteric, subtrochanteric, subcapital) and post-treatment outcomes (e.g., pain, mobility, mortality)?
• Key Question 3 – What is the relationship between implant variables (e.g., position, material, method, and design of implant) and patient post-treatment outcomes (e.g., pain, mobility, mortality)?
• Key Question 4 – What is the relationship between the type of intervention (e.g., internal fixation versus arthroplasty) and patient post-treatment outcomes (e.g., pain, mobility, mortality)?

Figure 5 graphically shows the linkages between the key questions. Key question 1 examines the larger question of how patient characteristics, including fracture type, relate to patient outcomes, while key question 2 focuses specifically on the relationship between fracture type and outcome, holding patient characteristics steady. Key questions 3 and 4 are differentiated by the level of analysis of device type, either at the level of specific device or whole classes of devices.

Figure 5

Linkages between key questions.

As shown earlier in this chapter, the device chosen to treat a hip fracture is highly reliant on the characteristics of the fracture in question. With fracture type and treatment choice so strongly correlated, the ability to tease apart the separate main effects of fracture type and surgical treatment on patient outcomes would require a more comprehensive model. While the key questions rely on basically the same analyses, the information to support them comes from two different literatures. Unfortunately, few studies provide detailed information on both patient characteristics and the procedure characteristics, and fewer include both sets of information in their analyses.

Therefore, a more comprehensive map of the elements which contribute to outcomes was developed and reviewed with the Technical Expert Panel (TEP) members (Appendix B).

Conceptual Model

Figure 6 maps the major factors which impact the final outcomes hip fracture patients experience. These outcomes are influenced by patient characteristics, the specific characteristics of the sustained fracture, and elements of the treatment, including effects related to the hospital setting, the device, the skills and abilities of the surgeon, the operative processes, and post-acute treatment rehabilitation. The shaded boxes in the model represent variables directly related to the key questions and are the focus of the trials included in this review. The white boxes represent the other factors which must also be accounted for in order to isolate the main effects of the fracture type and treatment choice and are contained within this review only to the extent that they are relevant covariants. The bolded arrows are those relationships related to the key questions.

Figure 6

Conceptual model.

Patient factors important to hip fractures include demographic, medical, and other risk factors predictive of patient outcomes.^4,6,51 Many of these factors enter a surgeon’s decision process when choosing specific device treatments, including the American Society of Anesthesiologists (ASA) score, used to recognize a patient’s health status and presence of comorbidities. Characteristics that have been found to impact patient outcomes include age, gender, cognitive status, social support, marital status, pre-fracture function, and pre-fracture residence.^13,76–78 One would anticipate that studies would capture these patient factors at baseline in order to ascertain the comparability of patients with regard to patient characteristics that influence patient outcomes and the appropriateness of device choices.

As discussed previously in this chapter, types of hip fractures include femoral neck, intertrochanteric, and subtrochanteric fractures. The fractures are further classified as displaced or nondisplaced among femoral neck fractures, and stable or unstable for pertrochanteric fractures. Fracture types and patterns are often not well documented due to the multiple classification systems used by surgeons to identify fracture patterns and subtypes. Studies need to provide clear, reliable, and consistent reporting of fracture patterns in order to establish the relationship between fracture patterns and outcomes.

Studies should additionally include a full complement of non-patient factors that potentially impact outcomes in each analysis. Some covariates may be controlled for through a randomized controlled study design, but the ability to draw inferences from pooled studies is difficult without the full set of covariates that extends beyond patient factors alone. Hospital settings, surgeon factors, operative factors, recovery and rehabilitation factors all impact hip fracture outcomes and should all be accounted for as much as possible within clinical studies.

Hospital factors which can influence a patient’s outcome trajectory include items such as annual volume of patients, specifically hip fracture patients, whether the hospital is a community hospital, teaching hospital or major trauma center, hospital location such as rural or urban, and the type of hospital ownership.⁷⁹

A number of operative factors may relate to surgical outcomes, such as delay to surgical treatment,⁸⁰ need for blood transfusions,⁸¹ and the duration of surgery.

Processes of care also differ by hospital location. Comprehensive processes of care can include prophylactic antibiotics and anticoagulants, urinary catheter protocols, mobilization and physical therapy, pain management, restraint protocols, and discharge processes for patients with active clinical issues.⁷⁷ Siu et al,⁷⁷ suggest that the benefit of any one intervention, such as thromboprophylaxis, is small or short-lived without timely followup with other elements of care, such as early weight-bearing. They found that combinations of processes of care were more predictive of patient functional outcomes than single interventions alone. However, the processes of care associated with lower mortality are not necessarily the same as processes of care associated with improved patient functionality.⁸²

Treatment for hip fractures does not end with hospitalization. Patients typically receive post-acute rehabilitative services for several weeks beyond the initial hospitalization. These services may be provided by skilled nursing facilities, acute rehabilitation units, home health care programs, or some combination of the above.^83–85 Given the presence of home health care programs, using data from patient discharge alone is not a good proxy for the full rehabilitative processes a patient undergoes.

The patient outcomes assessed in clinical studies should be aligned with the stated treatment goal of returning patients to their pre-fracture functional status. Pain, mobility, and the use of walking aids are common outcomes but are difficult to fully interpret without comparison to baseline abilities.³⁴ Magaziner et al.³⁴ found that recuperation times were specific to the type of function regained, ranging from 4 months to 1 year.

The conceptual model shown in Figure 6 guided the data abstraction and analysis for this review. In addition to reviewing particular study results, we also collected data regarding the presence or absence of classes of covariates that were used in order to assess the general validity of the research findings as a whole.

Scope of the Review

The review is focused on the patient population most likely to experience low energy traumatic fractures of the hip and the final patient outcomes of mortality, functional status, and quality of life. Patients younger than 50 and fractures due to high energy trauma are outside the scope of this review.

Many important factors noted in Figure 6 were not considered in this review. For example, we did not look at the direct effects of time delay to surgery, peri-operative monitoring, comprehensive perioperative medical management, or post-operative protocols. Attention to these items in this review was limited to examining whether such factors were accounted for when a study was looking for direct effects of surgical implants.

The review does not include analysis of devices at the level of specific manufacturers. The literature does not include sufficient information on manufacturers. Further, external fixators, devices that are affixed to the outside of the leg, or devices or surgical methods that are no longer in common use within the United States are outside the scope of this review. We also did not delve into broad classes of arthroplasty implants. For example, much of the literature on the topic of cemented versus cementless arthroplasty compares old technology uncemented implants (that would not be used for elective total hip replacement) to various cemented designs where technological advances have been less dramatic than for cementless designs.