Patient-reported physical functioning and quality of life after pelvic ring injury: A systematic review of the literature

Background Pelvic ring injuries are one of the most serious traumatic injuries with large consequences for the patients’ daily life. During recent years, the importance of the patients’ perception of their functioning and quality of life following injury has increasingly received attention. This systematic review reports on self-reported physical functioning and quality of life after all types of pelvic ring injuries. Methods The online databases MEDLINE-PubMed and Ovid-EMBASE were searched for studies published between 2008 and 2019 to identify published evidence of patient-reported physical functioning and quality of life after which they were assessed for their methodological quality. Results Of the 2577 articles, 46 were reviewed in full-text, including 3049 patients. Most studies were heterogeneous, with small cohorts of patients, a variety of injury types, treatment methods and use of different, often non-validated, outcome measures. The overall methodological quality was moderate to poor. Nine different PROMs were used, of which the Majeed Pelvic Score (MPS), SF-36 and EQ-5D were the most widely used. Mean scores respectively ranged from 75–95 (MPS), 53–69 (SF-36, physical functioning) and 0.63–0.80 (EQ-5D). Conclusions Physical functioning and quality of life following pelvic ring injuries seem fair and tend to improve during follow-up. However, differences in patient numbers, injury definition, treatment strategy, follow-up duration and type of PROMs used between studies hampers to elucidate the actual effects of pelvic ring injuries on a patient’s life. Implications of key findings Physicians and researchers should use valid and reliable patient-reported outcome instruments on large cohorts of patients with properly defined injuries to truly evaluate physical functioning and quality of life after pelvic ring injuries. Systematic review registration number PROSPERO International prospective register of systematic reviews; registration number CRD42019129176.

treatment strategy, follow-up duration and type of PROMs used between studies hampers to elucidate the actual effects of pelvic ring injuries on a patient's life.

Introduction
Pelvic ring injuries can be seen as one of the most serious traumatic injuries with large consequences for the patients' daily life. Apart from the substantial mortality rates [1,2], principally in high-energy trauma, these injuries coincide with long periods of impaired mobilization and intense rehabilitation. In addition, pelvic ring injuries are increasingly caused by low-energy trauma in the frail elderly. Injury types vary from stable type A fractures, usually treated nonoperatively, to highly unstable type C fractures, often demanding operative fixation and long term recovery. Despite this, adequate prospective follow-up studies, both on short-term and long-term outcome, on pelvic ring injuries are lacking.
Many factors that characterise a patient's health status cannot be observed, measured with a device, or analysed with even the most sophisticated imaging methods. How a patient feels and performs remains largely impenetrable to devices [3]. The growing focus on patient-centred care has resulted in a shift in terms of outcome assessment and the increasing use of Patient-Reported Outcome Measurements (PROMs). These questionnaires seek to assess the influence of the patients' condition on their daily functioning and emotional status, and can provide critical information to enhance patient-centred health care [4]. Conceptually, PROMs can be viewed either as a 'tool for evaluation' or as a 'mechanism for improvement'.
No actual guidance exists for appropriate PROM-based assessment after pelvic ring injuries. Hence, the problem arises with regard to the long list of different PROMs used, many of which have no proof of being valid or reliable either. Lefaivre et al. [5] showed that many different types of generic outcome instruments as well as pelvis-specific measures are used to assess the outcomes after pelvic ring injuries. Besides, due to the wide variety in types of pelvic ring injuries and the variability in treatment strategy, outcomes are hard to compare, leaving physicians, researchers and patients in doubt about the actual outcomes following these injuries.
In this perspective, the main objective of the present systematic review was to identify and analyse published studies, thereby providing a representative overview of the outcomes in terms of patient-reported physical functioning and quality of life following pelvic ring injuries. Moreover, following the results of this review, our aim was to highlight whether changes can be made for future research in order to properly evaluate the consequences of these severe injuries.

Methods
For this systematic review the PRISMA method [6] for literature collection and manuscript construction was followed. The review protocol has been registered in PROSPERO International prospective register of systematic reviews under registration number CRD42019129176.

Identification of studies: Search strategy
The search strategy sought to retrieve references relating to physical functioning and quality of life after pelvic ring injuries. Therefore, the items "pelvis", "injury" and "outcome" were combined to develop the search strategy. Searches used medical subject headings (MeSH terms) and free text searching to combine terms specific to pelvic ring injuries with terms relevant to PROMs evaluation. The full electronic search strategy was developed in collaboration with an experienced medical librarian and is presented in Table 1

Inclusion and exclusion criteria and procedure
Eligible studies included patients aged 18 years or older with a pelvic ring injury. Studies that focused on the outcomes after nonoperative as well as operative treatment were eligible. The outcome measures used should include patient-reported outcome measures (PROMs). Except for case studies and conference abstracts, all study designs were accepted for inclusion. Concerning language, studies written in English, German, Spanish, French and Dutch were included. There was no limitation on the search by publication status. Studies on geriatric fractures or fragility fractures were excluded. Studies with a sample size of less than 20 patients in follow-up were excluded, because PROMs results based on so few patients seem unreliable. Moreover, studies that included outcomes after both pelvic ring injuries and acetabular fractures and that did not differentiate between these injuries in terms of outcomes, were excluded as well. The study selection was performed in two screening phases: 1) title and abstract  The results of the various questionnaires are shown according to the standards of the specific  questionnaire, either as number with percentage or as mean with standard deviation or  median with range or interquartile range (IQR). Pooled means and standard deviations were manually calculated for the complete cohort of every study in case the outcomes of the PROMs were provided for two or more groups.

Selection of studies
The initial searches (conducted from January 2008 to April 15 th 2019) generated 2577 articles. Following title and abstract assessment, 95 articles were reviewed in full text. A total of 46 articles were included in the review, of which most (N = 22) were cross-sectional studies, followed by case-control studies (N = 12), cohort studies (N = 10), one RCT and one combination of a cohort and cross-sectional study. Fig 1 demonstrates a flowchart of the inclusion procedure.

Patient and injury characteristics
Overall, data of a total of 3049 patients were reported in the studies. The number of patients included in the studies varied widely, from as little as 20 patients [8] up to as much as 263 patients [9]. However, most studies were relatively small; only seven studies [9][10][11][12][13][14][15] included more than 100 patients and more than half reported on even less than 50 patients. Thirty-eight studies focused on unstable pelvic ring injuries (Type B and/or Type C according to the AO classification system [16]), whereas only six studies included all types of pelvic ring injuries [10,12,15,[17][18][19]. Two studies focused on the outcomes after sacral fractures [20,21]. Both nonoperative treatment as well as several operative techniques were applied to treat the patients, although no study solely focused on the outcomes after nonoperative treatment. Operative techniques varied from external fixation to internal fixation with osteosynthesis plates to percutaneous fixation and other minimally invasive techniques. All included studies are described in Table 2.

Methodological quality assessment
The results of the quality assessment of the included articles are presented in Table 3. Total scores in percentages ranged between 50% and 92%. The average score was 72%. No studies were excluded based on this assessment. Most studies scored fairly positive on the first four areas, regarding citation (1), study purpose (2), relevant background literature (3), and description of the sample (4). None of the studies justified sample size (5), which is the reason that no studies scored the maximum amount of points on the assessment. In the RCT [50] randomization of groups was performed (6), but it was not clearly described by which method (7). The first eight studies used valid (8) and reliable (9) PROMs, though some used both valid PROMs and PROMs of which the validity was not established (+/-). The ten studies in the list with the lowest quality scores did often not report results in terms of statistical significance (10) and did not use appropriate analysis methods (11). The last three areas regarding clinical importance (12), dropouts (13) and appropriate conclusions (14) were mostly sufficiently described.

Discussion
The management of and recovery of pelvic ring injuries has had gained attention over the years by clinicians and researchers. Although the focus primary laid on radiographic outcomes over the past decades, more recently this focus shifted towards the use of patient-reported outcome measures (PROMs). This is the first systematic review to evaluate outcomes in terms of physical functioning and quality of life after pelvic ring injuries. The extensive literature search resulted in the inclusion of 46 studies regarding patients with a broad range of injury types and treatment methods. Physical functioning and quality of life was mainly assessed between one and five years after pelvic ring injury. Most studies had small sample sizes, with more than half including even less than 50 patients. Besides, the quality of the studies was moderate to poor. Nine different outcome measures were used; 38 studies used disease-specific PROMs and 15 studies used generic PROMs. None of the disease-specific PROMs have been proven valid for use in patients with pelvic ring injuries. Overall, the recovery of physical functioning and quality of life following pelvic ring injuries seemed fair, although the reported results varied widely between studies and the different PROMs. Taking all of the above into account, it is challenging to conclude an overall result in terms of physical functioning and quality of life after pelvic ring injuries. Hence, some critical remarks can be made on the included studies based on the results of this systematic review.
Most studies reported on a wide variety of pelvic ring injury types. According to the AO/ OTA classification system [16], pelvic ring injuries can be divided into type A, B or C injuries. However, sometimes the Young-Burgess classification [59] was used, which divides these injuries into 'anterior posterior compression (APC)', 'lateral compression (LC)' or 'vertical shear injuries (VS)'. In the studies that were included in this systematic review, it was not always clear what type of injury the patients had and most studies did not differentiate in the outcomes between for example B and C type injuries. Although type B as well as type C injuries are considered to be unstable fractures, type B injuries are simply rotationally unstable and therefore more likely to result in good outcomes, compared to the rotationally as well as vertically unstable type C injuries. Also, type A injuries were only assessed in six studies [10,12,15,[17][18][19] even though this type consists most of all types of pelvic ring injuries [12]. Moreover, there was no differentiation in outcomes of patients with solely a pelvic ring injury,   and of patients with multiple injuries, which is seen in polytrauma patients. This may clearly affect results of generic PROMs. None of the studies focused solely on the outcomes after nonoperative treatment of pelvic ring injuries. Only a few of the included studies [10,12,53,57,60,[13][14][15]17,18,33,35,46] evaluated outcomes of patients that were treated either operatively or nonoperatively, while most studies only assessed operatively treated patients. Moreover, among the operatively treated patients, a wide variety of surgical techniques was used. The used techniques varied from external fixation, to purely anterior or posterior fixation, to a combination of both and even experimental techniques for specific pelvic ring injury types. Due to this variety in applied surgical techniques, which were often also poorly described, it was not possible to perform subgroup analyses. After all, the aim of this systematic review was to provide a general assessment of outcomes after pelvic ring injuries, but not of any specific operative approach.
Follow-up was mainly assessed between one and five years, missing the important shortterm (<12 months) as well as long-term (>5 years) consequences of these injuries on the patients' daily life. Especially in the studies evaluating surgical techniques, the short-term follow-up is highly important, as this is a critical period in which the most improvement in physical functioning can be achieved. On the other hand, long-term follow-up might be just as important, revealing the late complications like gait impairment, chronic pelvic and back pain as well as delayed consequences of lumbosacral plexus injury [61]. Also, the unknown preinjury condition for physical functioning and quality of life leaves us guessing about the actual effect of the injury on the patients daily life.
Another problem in the evaluation of the studies was that the sample sizes of most studies were small, often including even less than 50 patients (N = 24). The methodological quality assessment revealed that no sample size calculation was performed in each of the studies, which makes it arguable whether enough patients were included to draw conclusions from in terms of physical functioning and quality of life. The quality assessment also revealed that, overall, the methodological quality was moderate and did not reach perfection in any of the studies, as all missed the justification for sample size. Moreover, many studies failed to achieve higher scores due to the use of nonvalidated outcome measures like the MPS.
The use of nine different PROMs was another issue. Of the four different disease-specific PROMs, the MPS was by far the most frequently used PROM in 34 studies, even though it has never been validated in patients with pelvic ring injuries. The reason for its frequent use could be explained by the compact length of the questionnaire and the possibility to compare outcomes to those of other studies. Similar to the results of this review, Lefaivre et al. showed that the MPS is the most commonly used pelvic outcome score [5]. Results were most often graded as 'excellent', although there was a wide variation in the proportion of patients that had an excellent score between the various studies. Only three studies [12,13,57] used two different generic PROMs (MFA and SMFA) to assess physical functioning, while quality of life was assessed in 13 different studies using the SF-12, SF-36 and EQ-5D, showing acceptable quality of life following pelvic ring injuries. The asset of these generic questionnaires is the availability of normative data to compare results with. A complicating factor was that the scores on identical questionnaires were often reported in different ways, making them hardly comparable. For example, the results on the MPS of the SF-36 were frequently reported by the categories (excellent, good etcetera), whereas other studies only presented mean scores with standard deviation, range, or a combination of these. In addition, scores varied widely, even between studies that used the same PROMs.
None of the disease-specific questionnaires that were used have been proven to be valid to assess physical functioning of patients with pelvic ring injuries, while all generic outcome instruments have. The ability of the outcomes of PROMs to improve decision-making in clinical research relies on the psychometric strength of the instrument to capture the burden of disease or treatment. Reliability and validity are separate psychometric properties, both essential for any measure [62]. Measures can be highly reliable but not measure what they are supposed to measure [63]. Some studies compared pelvic-specific PROMs with generic PROMs to investigate the validity of disease-specific instruments in examining pelvic-specific areas, but failed to do so [5,[64][65][66]. Hence, until there is a disease-specific questionnaire for pelvic ring injuries that is proven to be valid and reliable, it seems preferable to use a reliable and valid generic PROM to assess physical functioning and quality of life following these injuries. Another advantage of the latter is that, for these generic PROMs normative data often available is.
PROMs enable important clinical questions to be answered in clinical research [3]. Its use should be integrated in the clinical evaluation of a patient with pelvic ring injuries, next to the more objective measures like radiographic outcomes, because PROMs directly reflect the patients' perspective on the impact of their injury on daily life. Some types of pelvic injuries may look highly unfavourable on radiographic imaging, but the patient may grade his physical functioning and quality of life fairly well, or the other way around. Despite the fact that there has been discussion on the actual contribution of PROMs to the improvement of patient care, these instruments have the potential to facilitate patient involvement in treatment decision-making and provide guidance for health-care decisions [63]. Patients may monitor their health status over time and eventually will be more actively engaged in striving for health outcomes like full rehabilitation. Also, PROMs may help clinicians quickly identify which of their patients experience improved or deteriorated health outcomes. This may help to identify any structural patient complaints, which would suggest that refinements to care pathways might be needed. However, at this moment, PROMs function more as a tool for the use in clinical research, than they do in substantially changing medical practice.

Strengths and limitations
Some strengths and limitations of this systematic review and its conclusions need to be addressed. To start with, this is the first systematic review to evaluate patient-reported physical functioning and quality of life after pelvic ring injuries. Also, search criteria were not limited by the type of study (e.g. cohort study, RCT), which provided a complete overview of all study results published during the past decade. Moreover, this systematic review underlines that some changes are needed in the future in order to examine the true consequences of pelvic ring injuries on the patients' daily life, for example to only use reliable and valid patient-reported outcome instruments. In this systematic review, a highly sensitive comprehensive search was conducted following the recommendations of an experienced medical librarian in order to identify articles of interest. For practical reasons though, only studies published in English, German, Spanish, French or Dutch were included in the final review, which might have led to selection bias. Additionally, studies published before 2008 were excluded after consultation with two experienced pelvic trauma surgeons. The argument for this was that, before 2008, treatment methods differed such an extent that including studies published before that time might lead to bias in the results of this systematic review. In this review, we included all types of pelvic ring injuries, treatment methods and types of PROMs. Due to this heterogeneity, individual outcomes of the included studies were not suitable for reliable comparisons. At last, sample sizes were not justified in any of the included studies.

Conclusion
Even though the above-mentioned critical remarks make it ambitious to draw conclusions in terms of physical functioning and quality of life after pelvic ring injuries, the results imply that patients' physical functioning and quality of life seem reasonably fair and improve over time. However, a heterogeneous group of studies was presented, including small cohorts of patients with a wide range of injury types, treatment methods and diverse, often nonvalidated, outcome measures. Hence, there is a high need to use a valid and reliable outcome measure to evaluate and compare the recovery in terms of physical functioning and quality of life after pelvic ring injuries on large groups of patients. The following section provides some guidance for future research.

Practical implications and recommendations for future research regarding use of PROMs after pelvic ring injuries
• Authors should clearly define the injury type according to the AO/OTA classification and distinguish between outcomes of different types of injuries. They should also distinguish between a pelvic ring injury as the only injury or as part of multiple injuries.
• Prospective longitudinal studies are needed with sufficient number of patients and multiple time intervals at short-term as well as long-term (>5 years) follow-up.
• (Recalled) pre-injury status of physical functioning and quality of life should be recorded.
• Only valid and reliable PROMs should be used, for example the SMFA for physical functioning and the EQ-5D or SF-36 for quality of life. These PROMs can be compared with age-specific norm data of the general population. The use of non-validated pelvic-specific PROMs should be avoided.
• There is still a challenging and a necessary task to validate existing pelvic-specific PROMs and develop an uniform PROM for pelvic injuries worldwide.