The Use of Mri in the Diagnosis of Occult Hip Fractures in the Elderly a Preliminary Review

Proximal femoral fractures are a frequent cause of morbidity and mortality in elderly patients with osteoporosis [ane]. In patients with acute hip pain and suspected fracture after a fall or minor trauma, the American College of Radiology (ACR) Appropriateness Criteria recommend initial imaging with conventional radiography [ii]. ACR modified the guidelines in 2019 to advocate for either CT or MRI as "unremarkably appropriate" if radio-graphic findings are negative for a fracture just a fracture remains suspected. This recommendation represents a change from the previous 2014 ACR guidelines in which MRI was accounted "usually appropriate" and CT was labeled every bit "may exist appropriate," an improvement of at least 1 betoken in the CT appropriateness rating [three]. In office, this change is because of contempo studies showing improved accuracy with CT. A 2019 meta-analysis showed a sensitivity and specificity of CT for radiographically occult proximal hip fractures of 94% and 100%, respectively [4]. An additional do good of CT is speed of imaging, particularly for patients with associated confusion at the time of injury [2]. The hateful scanning fourth dimension for some multiparametric MRI hip fracture protocols can exist up to 30 minutes, which is not applied for patients with difficulties following instructions [v].

In contrast to a 2019 meta-analysis that reported that CT yielded false-negative results in up to 10% (50/496) of hip fractures [4], early studies suggested near-perfect functioning of MRI for the detection of hip fractures [6, 7]; withal, these studies were methodologically flawed [vi, 7]. To reduce the brake of MRI in this setting considering of time requirements, limited MRI protocols have been developed. One report of patients with hip hurting showed a 100% accuracy of a limited coronal STIR imaging protocol for the detection of various abnormalities compared with multiparametric MRI [viii]. The performance of limited MRI protocols for the detection of proximal femoral fractures needs to be evaluated systematically to determine the quality and generalizability of bachelor data to date. The primary objective of this systematic review and meta-assay was to evaluate the diagnostic accurateness of limited MRI protocols for detecting radio-graphically occult proximal femoral fractures compared with a multiparametric MRI protocol with or without clinical issue equally the reference standard. Primary secondary objectives included evaluating the diagnostic accuracy of unmarried-plane T1-weighted, STIR, T1-weighted and STIR, and T2-weighted protocols for the detection of radiographically occult proximal femoral fractures. Other secondary objectives included evaluating agreement betwixt reviewers, evaluating scanning fourth dimension reduction with limited MRI protocols compared with multiparametric MRI protocols, and evaluating for other possible causes of variability including MRI magnet strength and reviewer training with subgroup analyses.

This systematic review and meta-analysis was conducted and reported in accordance with electric current best practices and the Preferred Reporting Items for Systematic Reviews and Meta-Analyses diagnostic exam accuracy (PRISMA-DTA) guideline [ix–eleven]. The review is based on an a priori established protocol that was registered in the International Prospective Register of Systematic Reviews (PROSPERO) database before initiation (registration number CRD42020158731). This study was exempt from upstanding approval at the University of Alberta considering the analysis involved only de-identified data and because private studies had received prior ideals approving when required.

Literature Search

Comprehensive database searches of MEDLINE, Embase, Scopus, and the Cochrane Library (including the Cochrane Central Register of Controlled Trials, Cochrane Central Annals of Protocols, and the Cochrane Database of Systematic Reviews) through November 15, 2019, were performed to identify studies that used an abbreviated MRI protocol for detecting radiographically occult hip fractures. A combination and variation of championship, abstract, and keywords and medical discipline heading terms were used including MR* AND fracture AND hip AND (sensitivity OR specificity OR accuracy). The specific search criteria by database are shown in Appendix 1. No beginning engagement or linguistic communication restrictions were applied, and translation was acquired when necessary. The search was performed according to all-time practices for electronic search strategies [12].

Manufactures from each database were so combined, and duplicate manufactures were removed from the listing. Titles and abstracts were screened for relevance, and the total text of all relevant articles was retrieved to evaluate for satisfaction of the inclusion criteria. The reference lists of the relevant studies were checked. Forward searching of relevant articles was performed in Google Scholar. The systematic review of databases, including title and abstract review and full text review, was performed independently by two reviewers (reviewers ane and 2). Consensus was used when determining manufactures selected for full-text review likewise as selecting full-text articles sufficient for inclusion. A search of the gray literature was likewise performed by one writer (reviewer 2) who evaluated the most recent three years (2017–2019) of major radiologic and orthopedic conference proceedings including the Radiological Society of Northward America, the International Society for Magnetic Resonance in Medicine, the Society of Skeletal Radiology, and the Orthopaedic Trauma Association. The most recent coming together (2019) of the ARRS was also reviewed. Briefing abstracts with sufficient information to meet the inclusion criteria were included in the analysis.. Conference abstracts for which subsequently published articles were bachelor for review were not included in farther cess.

Selection Criteria

All original articles evaluating the diagnostic accuracy of a express MRI protocol for detection of radiographically occult proximal femoral fractures in patients with acute hip hurting after minor trauma compared with a multiparametric MRI protocol with or without clinical outcome as the reference standard were evaluated with full-text review. Nosotros required that studies take sufficient data to reconstruct a 2 × 2 contingency tabular array. If information in an article was not sufficient to reconstruct a 2 × 2 contingency table, we contacted the authors via due east-mail to obtain the missing information; if sufficient information was supplied, the study was included in the assay.

Based on an a priori protocol, studies were excluded from the analysis if just pediatric patients younger than 18 years old were evaluated, fewer than 10 patients were included, a patient population other than patients with acute hip hurting after trauma was used, fractures that were not radio-graphically occult were included, a limited MRI protocol was not the index test, multiparametric MRI was not the reference standard, or there was insufficient information available to reconstruct a 2 × ii contingency tabular array later on attempted e-mail correspondence with the author or authors. In addition, nonoriginal manufactures including review articles, guidelines, consensus statements, letters, and editorials were excluded.

Information Extraction

The relevant data from included studies were extracted independently past two reviewers (reviewers 1 and 2). Patient characteristics including the following were recorded: total number of patients, number of patients with hip fractures, mean patient age, and sex. Study characteristics including the following were recorded: first author's name, twelvemonth of publication, clinical setting (bookish vs private), prospective or retrospective study design, single-heart or multicenter information conquering, number of readers, review strategy (independent vs consensus), and level of preparation of the reader or readers. Details regarding MRI characteristics including model type, magnet force, sequence or sequences used in the abbreviated and multipara-metric protocols, TRs and TEs of the predominant abbreviated sequence, FOV, repeat-train length, slice thickness, matrix size, number of signals acquired, and full scanning times of abbreviated and multiparametric protocols were recorded. If available, reported truthful-positive (TP), false-negative (FN), truthful-negative (TN), and faux-positive (FP) results were recorded. If TP, FN, TN, and FP results were not reported, fracture prevalence, sensitivity, specificity, and accuracy or AUC values were recorded, and a 2 × 2 tabular array was reconstructed using the reported sample size, prevalence, sensitivity, specificity, and accuracy or AUC values. For studies with multiple reviewers, results from each reviewer were recorded and analyzed separately [13]. In areas in which data were inconsistent betwixt the reviewers, the commodity was reviewed and consensus was reached. If disagreement persisted, the data were reviewed by a third author (reviewer 3). Data were collected and recorded in Excel (version 15.fourteen, Microsoft).

Adventure of Bias Assessment

The methodologic quality of each written report was evaluated using the Quality Assessment of Diagnostic Accurateness Studies–2 (QUADAS-2) tool [14]. The QUADAS-2 tool is used in systematic reviews to review private studies for potential sources of bias by evaluating the following 4 domains: patient choice, index examination or tests, reference standard, and menses and timing. The QUADAS-ii scores for each domain were independently assessed past ii reviewers (reviewers 1 and two). No discrepancies were present betwixt reviewers.

Data Analysis

A 2 × ii contingency table was created for each study. Meta-analysis was performed using a bivariate mixed-effects regression model. The model fits 2 levels: independent binomial distributions for the TPs and TNs conditional on the sensitivity and specificity in each study and a bivariate normal model for the logit transforms of sensitivity and specificity between studies [15, 16]. From the extracted data, summary sensitivities, specifici-ties, both negative and positive likelihood ratios, and diagnostic odds ratio (DOR) with corresponding 95% CIs were calculated to evaluate the operation of abbreviated MRI protocols. In improver, the surface area under the summary ROC (SROC) curve was calculated every bit an alternative global mensurate of accuracy to reduce the influence of heterogeneity.

Threshold event was evaluated using a ROC bend for abbreviated MRI protocol functioning. If the threshold effect exists, a curvilinear pattern should exist present. Interstudy variability is not reported with Q or I ² statistics, consistency tests that are routinely performed in intervention reviews, considering these statistics are no longer supported past the PRISMA-DTA Group [x]. Variability was assumed, and sources for variability were examined if possible. A Deek funnel plot was also not performed because it is no longer considered an constructive means to measure publication bias according to the PRISMA-DTA Grouping [x]. Analysis was conducted using Stata software (version 16, StataCorp).

Literature Search

The literature search was conducted according to PRISMA guidelines and is shown in Figure 1 . Of the 7 studies identified on the Cochrane Library database, six studies were from the Cochrane Key Register of Controlled Trials and one study was from the Cochrane Database of Systematic Reviews. Eleven studies met the inclusion criteria, with all studies providing sufficient information to construct a two × 2 contingency table [7, 17–26]. Vi of these studies used only a express MRI protocol with a reference standard of the same protocol in combination with clinical outcome with or without corresponding CT [7, 17–21]. These studies were excluded from the meta-assay. 5 studies evaluated a express MRI protocol compared with a multiparametric MRI protocol and were included in the meta-analysis [22–26]. Two of these studies evaluated a mix of femur, pelvic, and sacral fractures, only subgroup information for merely femoral fractures were developed after contacting the authors [22, 24].

View larger version (57K)

Fig. 1 —Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow diagram shows screening and pick of studies included in systematic review.

Patient, Study, and MRI Characteristics

A full of 938 patients with 247 proximal femoral fractures were included in this review. The mean age of the pooled cohort was seventy years (range, 12–100 years), and 69% (575/838) of the population was female in studies reporting sex. Of the 938 patients, 602 patients with 110 proximal femoral fractures were included in the meta-analysis. One study of 22 patients evaluated each hip for each patient separately as segmented studies, totalling 44 reviewed hips with three proximal femoral fractures [22].

Study characteristics of individual included studies are detailed in Table 1. All studies were performed at single bookish centers. The study pattern was prospective for five studies and retrospective for half-dozen studies. Of the five studies included in the meta-analysis, four used multiple readers reporting separately and one written report used a single consequence afterward consensus reading [25]. Reviewers were a mix of radiology residents, fellowship-trained radiologists, and orthopedic surgeons. Most studies used an MRI scanner manufactured by Siemens Healthineers with a mix of magnet strengths. Technical parameters for study sequences are listed in Table ii. Three studies reported mean scanning times for limited and multiparametric protocols with mean reported times of 4 minutes 49 seconds and at least 13 minutes 41 seconds, respectively [seven, 22, 26]. Scanning fourth dimension for multiparametric MRI is underestimated because i written report just reported scanning fourth dimension for the minimum number of included sequences rather than mean time for all multisequence studies [22].

TABLE 1: Characteristics of Studies Included in and Studies Excluded From Meta-Analysis

Table 2: MRI Protocols by Sequence for Studies Included in and Excluded From the Meta-Analysis

Diagnostic Accuracy of Limited MRI for Hip Fracture

Individual written report results with a list of corresponding 2 × 2 contingency table data and reader agreement (kappa statistic) are shown in Table three. Sensitivity and specificity of limited protocols for detecting radiographically occult proximal femoral fractures ranged between 84% and 100% and between 95% and 100%, respectively. The pooled and weighted summary sensitivity and specificity values were 99% (95% CI, 91–100%) and 99% (95% CI, 97–100%), respectively (Fig. two ). The area under the SROC curve was 1 (95% CI, 0.99–1). The positive likelihood ratio, negative likelihood ratio, and DOR were 89 (95% CI, 38–208), 0.01 (95% CI, 0–0.one), and 7231 (95% CI, 488–107,228), respectively. Studies evaluating interobserver agreement amongst reviewers included kappa statistic values entirely within the almost perfect to perfect range (Table three).

Tabular array three: Information for 2 × 2 Contingency Tables Evaluating Express MRI Protocols for Detecting Radiographically Occult Hip Fractures by MRI Sequence or Sequences Used in Limited Protocol and Individual Reader in Studies With Contained Review

View larger version (85K)

Fig. 2 —Pooled sensitivity and specificity of limited MRI protocols for detecting radiographically occult hip fractures. Diamonds show conviction intervals, and dashed lines testify ways of pooled and weighted results. T1 = T1-weighted, T2 = T2-weighted, R1 = reader 1, R2 = reader 2, R3 = reader 3.

Express protocol type, MRI magnet strength, and level of reviewer training were evaluated. Only level of reviewer training could be evaluated with meta-analysis because of a small number of events in FP and FN categories for other groups. An abbreviated MRI protocol composed of merely a coronal STIR sequence showed an aggregate sensitivity of 99% (126/127), which improved to 100% with the improver of a T1-weighted sequence (118/118). A coronal T2-weighted sequence solitary had the lowest aggregate sensitivity at 86% (51/59), although performance was perfect in a study using T2-weighted Dixon sequences (9/ix). Aggregate sensitivity and specificity of a T1-weighted sequence simply, a STIR sequence only, T1-weighted and STIR sequences, and a T2-weighted sequence only were 97% (89/92) and 100% (76/76), 99% (126/127) and 99% (865/873), 100% (118/118) and 99% (867/874), and 86% (51/59) and 97% (137/141), respectively.

Studies using simply 3-T magnets showed a slightly improved aggregate sensitivity compared with studies using magnet strengths less than iii T (100% [58/58] vs 96% [326/338]), although specificity was like. Amass sensitivity and specificity of MRI protocols using only three-T units were 100% (58/58) and 97% (416/428), respectively, and aggregate sensitivity and specificity of MRI protocols using units with magnet strengths other than 3 T lone were 96% (326/338) and 99.5% (1529/1536). Studies using magnet strengths other than 3 T alone were predominantly composed of 1.5-T images but included a range of 1- to three-T imaging.

In subgroup analysis, radiologists showed a slightly improved pooled sensitivity compared with radiology residents and orthopedic surgeons (99% [95% CI, 90–100%] vs 95% [95% CI, 87–98%]), only this divergence was non statistically significant. No statistical differences were present in specificity or accuracy betwixt these groups. Four studies described the location of each proximal femoral fracture, but no studies provided sufficient data to perform a subgroup assay for intracapsular versus intertrochanteric location [23–26].

Gamble of Bias Assessment

Risk of bias cess is outlined in Appendix 1. All studies included in the meta-analysis were low gamble. The half-dozen studies included in the review but not in the meta-assay were high risk for both alphabetize test and reference standard because the studies evaluated for hip fractures merely with a limited MRI protocol and used the express protocol with or without clinical outcome outcome as the reference standard [vii, 17–21]. One study used a combination of MRI, clinical outcome, and CT as a reference standard [nineteen]. Given that the reviewers were non blinded to the reference standard or to the index test when interpreting the other, these studies were not included in the meta-analysis.

Reducing the time required for MRI has the potential to increase availability of MRI scanner fourth dimension in the acute setting while also reducing the likelihood of a failed exam in patients who are unable to remain however. The results of our meta-assay show near-perfect operation of express MRI protocols for the detection of radiographically occult proximal femoral fractures compared with a multiparametric MRI protocol with or without clinical outcome every bit the reference standard with a near-threefold reduction in active scanner time. This result persisted regardless of the specific protocol used; all the same, aggregate performance was best with the protocol of a STIR sequence only and the protocol of T1-weighted and STIR sequences. The sensitivity of MRI for proximal femoral fractures is higher than the sensitivity of CT recently reported in a meta-analysis reviewing CT performance in which an FN study occurred in as many as ten% of cases [iv]. In that study, Kellock et al. [iv] recommended MRI follow-up in patients with negative CT findings but ongoing pain. In addition to confidently diagnosing the presence of a proximal femoral fracture in the acute setting, the well-nigh-perfect sensitivity of a limited protocol MRI can allow prudent advancement of weight-begetting status while as well reducing the demand for further imaging follow-up, caution-ary infirmary admission, or skilled nursing facility placement in patients if appropriate out-patient follow-up is bachelor.

Reader feel has been previously shown to contribute to diagnostic accuracy of MRI for radiographically occult hip fractures with more senior readers showing 100% accurateness [27]. Although a slight comeback in sensitivity was identified with radiologist readers compared with radiology residents and orthopedic surgeons, the specificity and accurateness were not significantly unlike. When orthopedic surgeons are excluded from comparison, the difference is fifty-fifty less apparent. Additionally, slightly improved sensitivity is seen with studies using merely 3-T magnets, although this issue is also likely within the range of fault. Our evaluation suggests that functioning of limited protocols for proximal hip fractures is preserved despite differences in reviewer experience or magnet forcefulness.

This meta-analysis limited the diagnosis only to the presence of a proximal femoral fracture based on an a priori protocol. Previous studies accept noted the added benefit of MRI for diagnosing other pathologic findings including pelvic fractures and soft-tissue injuries [28–thirty]. The identification of nonsurgical fractures and soft-tissue injuries can contribute to appropriate alternative direction strategies such equally weight-bearing status, pain management, rehabilitation requirements, so on. The ability of limited MRI protocols to detect these findings has been previously reported in a study evaluating 93 MRI studies [8]; in that study [viii], two reviewers evaluated limited MRI sequences for the detection of underlying os, articulation, muscle, and bursal pathologic findings with 100% accuracy and with near-perfect agreement (κ = 0.96). Two studies with 5 reviewers included in our analysis also provided data evaluating the operation of limited MRI protocols for detecting any fractures of the proximal femur, pelvis, or sacrum [22, 24]. In these studies, the performance remained near perfect regardless of sequence type or reviewer preparation (accuracy, 95.3–99.9%). ACR appropriateness criteria recognize that unlike proximal femoral fractures, these other fractures and soft-tissue injuries oft do not necessitate urgent diagnosis [2]. They too notation an increasing trend toward conservative management of incomplete intertrochanteric fractures but maintain the utility of MRI for accurate fracture characterization and do not differentiate recommendations for subcapsular from intertrochanteric fractures. A subgroup analysis comparing these fracture patterns could not be performed in this meta-analysis.

Unlike most reviews of diagnostic examination accurateness, this study does non announced to be limited by meaning variability among studies [9–xi]. Despite this, we used a random-effects model that attempts to draw sensitivity and specificity with a mean and variance rather than estimate a setting-specific sensitivity and specificity as a means to mitigate the risk of variability. We as well performed subgroup analysis, when possible, to evaluate for potential sources of variability. This study as well is non affected by other frequent limitations of reviews of test accuracy, such equally small sample size. No current exam to sufficiently exclude the risk of publication bias from a review of examination accuracy is available, just we believe that this bias does not contribute significantly to our results [ten].

Although this systematic review notes a decrease in agile scanner time with use of a express MRI protocol compared with a multiparametric MRI protocol, this review does not specifically compare the difference in diagnostic paradigm quality of individual sequences in a limited MRI protocol versus those in a multiparametric MRI protocol and does not evaluate the entire time required to perform an MRI test (including transfer time and time between sequence acquisitions) in this patient population. The benefit of rapid and accurate diagnosis in patients with suspected hip fractures is well established and includes a mortality benefit, although the all-time imaging pathway to identify hip fractures is not withal established [31, 32]. The theoretic benefit of a limited protocol MRI has been proposed since the 1990s, although the specific comparative diagnostic, time, and price analyses of a limited protocol MRI test versus a CT test for radiographically occult hip fractures have not been evaluated systematically. These comparisons offering opportunities for futurity research.

This systematic review and meta-analysis shows most-perfect performance of limited MRI protocols for the diagnosis of radio-graphically occult proximal femoral fractures. The STIR sequence merely and the combination of T1-weighted and STIR sequences appear to testify the best performance. Limited MRI protocols can be used every bit the standard of care in patients with a suspected but radiographically occult hip fracture after trauma.

1. Brauer CA, Coca-Perraillon M, Cutler DM, Rosen AB. Incidence and mortality of hip fractures in the United states. JAMA 2009; 302:1573–1579 [Google Scholar]

2. Expert Panel on Musculoskeletal Imaging; Ross AB, Lee KS, Chang EY, et al. ACR Ceremoniousness Criteria acute hip pain: suspected fracture. J Am Coll Radiol 2019; 16(suppl 5):S18–S25 [Google Scholar]

three. Ward RJ, Weissman BN, Kransdorf MJ, et al. ACR Appropriateness Criteria acute hip pain: suspected fracture. J Am Coll Radiol 2014; 11:114–120 [Google Scholar]

iv. Kellock TT, Khurana B, Mandell JC. Diagnostic functioning of CT for occult proximal femoral fractures: a systematic review and meta-assay. AJR 2019; 213:1324–1330 [Abstract] [Google Scholar]

5. Ross A, Chan B, Yi P, et al. Rapid hip MRI: feasibility of detecting radiographically occult hip fractures in the elderly. (abstract) Skeletal Radiol 2017; 46:415 [Google Scholar]

six. Deutsch AL, Mink JH, Waxman AD. Occult fractures of the proximal femur: MR imaging. Radiology 1989; 170:113–116 [Google Scholar]

vii. Quinn SF, McCarthy JL. Prospective evaluation of patients with suspected hip fracture and indeterminate radiographs: employ of T1-weighted MR images. Radiology 1993; 187:469–471 [Google Scholar]

viii. Khoury NJ, Birjawi GA, Chaaya M, Hourani MH. Use of limited MR protocol (coronal STIR) in the evaluation of patients with hip pain. Skeletal Radiol 2003; 32:567–574 [Google Scholar]

10. McInnes MDF, Moher D, Thombs BD, et al. Preferred reporting items for a systematic review and meta-assay of diagnostic exam accurateness studies: the PRISMA-DTA statement. JAMA 2018; 319:388–396 [Google Scholar]

11. McGrath TA, Bossuyt PM, Cronin P, et al. All-time practices for MRI systematic reviews and meta-analyses. J Magn Reson Imaging 2019; 49:e51–e64 [Google Scholar]

12. Sampson 1000, McGowan J, Cogo E, Grimshaw J, Moher D, Lefebvre C. An bear witness-based practice guideline for the peer review of electronic search strategies. J Clin Epidemiol 2009; 62:944–952 [Google Scholar]

13. McGrath TA, McInnes MDF, Langer FW, Hong J, Korevaar DA, Bossuyt PMM. Treatment of multiple test readers in diagnostic accuracy systematic reviews-meta-analyses of imaging studies. Eur J Radiol 2017; 93:59–64 [Google Scholar]

14. Whiting PF, Rutjes AW, Westwood ME, et al.; QUADAS-2 Grouping. QUADAS-2: a revised tool for the quality cess of diagnostic accuracy studies. Ann Intern Med 2011; 155:529–536 [Google Scholar]

15. Reitsma JB, Glas AS, Rutjes AW, Scholten RJ, Bossuyt PM, Zwinderman AH. Bivariate analysis of sensitivity and specificity produces informative summary measures in diagnostic reviews. J Clin Epidemiol 2005; 58:982–990 [Google Scholar]

xvi. Riley RD, Abrams KR, Sutton AJ, Lambert PC, Thompson JR.

The benefits and limitations of multivariate meta-analysis, with application to diagnostic and prognostic studies

. Leicester, United kingdom of great britain and northern ireland: University of Leicester, 2005: technical report 05-04 [Google Scholar]

17. Bogost GA, Lizerbram EK, Crues JV 3rd. MR imaging in evaluation of suspected hip fracture: frequency of unsuspected os and soft-tissue injury. Radiology 1995; 197:263–267 [Google Scholar]

eighteen. Lim KB, Eng AK, Chng SM, Tan AG, Thoo FL, Low CO. Express magnetic resonance imaging (MRI) and the occult hip fracture. Ann Acad Med Singapore 2002; 31:607–610 [Google Scholar]

xix. Rehman H, Clement RG, Perks F, White TO. Imaging of occult hip fractures: CT or MRI? Injury 2016; 47:1297–1301 [Google Scholar]

xx. Rizzo PF, Gould ES, Lyden JP, Asnis SE. Diagnosis of occult fractures about the hip: magnetic resonance imaging compared with bone-scanning. J Bone Joint Surg Am 1993; 75:395–401 [Google Scholar]

21. Stiris MG, Lilleås FG. MR findings in cases of suspected impacted fracture of the femoral neck. Acta Radiol 1997; 38:863–866 [Google Scholar]

22. Heynen B, Tamigneaux C, Pasoglou V, Malghem J, Vande Berg B, Kirchgesner T. MRI detection of radiographically occult fractures of the hip and pelvis in the elderly: comparison of T2-weighted Dixon sequence with T1-weighted and STIR sequences. Diagn Interv Imaging 2019; 100:169–175 [Google Scholar]

23. Iwata T, Nozawa Due south, Dohjima T, et al. The value of T1-weighted coronal MRI scans in diagnosing occult fracture of the hip. J Bone Joint Surg Br 2012; 94:969–973 [Google Scholar]

24. Khurana B, Okanobo H, Ossiani M, Ledbetter S, Al Dulaimy K, Sodickson A. Abbreviated MRI for patients presenting to the emergency department with hip pain. AJR 2012; 198:[web]W581–W588 [Abstruse] [Google Scholar]

25. Lord C, Leach Due south, Skiadas VT. MRI for suspected hip fracture: how many sequences are enough? European Society of Musculoskeletal Radiology website. epos.myesr.org/poster/esr/essr2017/P-0268 Published 2017. Accessed December 1, 2019 [Google Scholar]

26. Ross AB, Chan By, Yi PH, Repplinger MD, Vanness DJ, Lee KS. Diagnostic accuracy of an abbreviated MRI protocol for detecting radio-graphically occult hip and pelvis fractures in the elderly. Skeletal Radiol 2019; 48:103–108 [Google Scholar]

27. Verbeeten KM, Hermann KL, Hasselqvist M, et al. The advantages of MRI in the detection of occult hip fractures. Eur Radiol 2005; 15:165–169 [Google Scholar]

28. Collin D, Geijer M, Göthlin JH. Prevalence of exclusively and concomitant pelvic fractures at magnetic resonance imaging of doubtable and occult hip fractures. Emerg Radiol 2016; 23:17–21 [Google Scholar]

29. Ohishi T, Ito T, Suzuki D, Banno T, Honda Y. Occult hip and pelvic fractures and accompanying muscle injuries effectually the hip. Curvation Orthop Trauma Surg 2012; 132:105–112 [Google Scholar]

30. Cabarrus MC, Ambekar A, Lu Y, Link TM. MRI and CT of insufficiency fractures of the pelvis and the proximal femur. AJR 2008; 191:995–1001 [Abstract] [Google Scholar]

31. Casaletto JA, Gatt R. Post-operative mortality related to waiting fourth dimension for hip fracture surgery. Injury 2004; 35:114–120 [Google Scholar]

32. Clague JE, Craddock E, Andrew One thousand, Horan MA, Pendleton N. Predictors of outcome following hip fracture: admission time predicts length of stay and in-hospital mortality. Injury 2002; 33:1–half dozen [Google Scholar]

33. Viera AJ, Garrett JM. Understanding interobserver agreement: the kappa statistic. Fam Med 2005; 37:360–363 [Google Scholar]

APPENDIX i: Search Strategy and Quality Cess of Diagnostic Accuracy Studies–2 (QUADAS-2) Assessment for Risk of Bias

myersshatuse.blogspot.com

Source: https://www.ajronline.org/doi/abs/10.2214/AJR.19.22676

Share this post