Abstract
Introduction:
Screening for osteoporosis is mainly done by measuring bone mineral density using dual-energy X-ray absorptiometry (DXA) scan. Screening does not usually start until 65 years. Many people 50 to 65 years may unknowingly suffer from progressive bone loss and may be at risk of osteoporosis. Opportunistic use of chest X-rays obtained for other indications could potentially be used to identify patients with signs of bone demineralization and at risk of osteoporosis in primary care settings particularly where DXA may not be readily available. We aimed at comparing the prevalence of bone demineralization using chest radiographs obtained for any indication in patients 50 to 65 years seen at an emergency department, as assessed from their radiology reports and by independent review by a radiologist. We also sought to explore clinical and demographic associates of low bone mass (LBM) among this population.
Methods:
We conducted a cross-sectional review of electronic medical records of patients 50 to 65 years, who had obtained a chest X-ray at the emergency department of a large Level 1 trauma center in an urban teaching hospital between May 1, 2021, and May 31, 2021. We estimated the prevalence of “bone demineralization” from the X-rays reports and compared it with that obtained by an independent radiologist’s review. We also used the 2020 US population census figures to estimate the number of people 50 to 65 years potentially at risk of LBM and investigated demographic and clinical correlates of LBM.
Results:
We reviewed 390 patient records, comprising 201 (51.5%) males and 189 (48.5%) females. Bone demineralization comment was found in 4 (1.0%) reports while radiologist review found 49 (12.6%). Applying sex-adjusted prevalence of 13.1% (using the direct method, with the 2020 US population as standard) to the 64.1 million adults 50 to 64 years in the US population, we estimated that approximately 8.4 million people will have bone demineralization in the population group. LBM was associated with hypertension in women (OR = 2.41, 95% CI = 1.03-5.64).
Conclusion:
Use of opportunistic chest X-rays may be feasible in identifying patients at risk of bone loss outside the traditional screening age particularly in areas where DXA may not be readily available.
Introduction
Osteoporosis is a skeletal disorder characterized by loss of bone mass and microarchitectural deterioration resulting in bone fragility and an increased risk of fracture. 1 Screening for osteoporosis is traditionally done by measuring the bone mineral density (BMD) using dual-energy X-ray absorptiometry (DXA) scan. Guidelines suggest screening for osteoporosis should begin at age 65 years in women, or earlier in those who have risk factors for fracture.1,2 While there is no screening recommendation for men by the United States Preventative Services Task Force (USPSTF), most recommendations suggest screening to begin at 70 years in those without any added risk. 1 However, many patients ranging from ages 50 to 65 years may unknowingly suffer from progressive bone loss until they experience a fragility fracture or consequent loss of physical function. Although fracture risk assessment by FRAX 3 is an accessible tool, it may fail to identify at risk patients because some risk factors observed by the primary care physicians (PCP) cannot be used in calculating fracture risk, for example, hypertension. 4 In addition, a PCP may not think about FRAX unless the visit is relevant to osteoporosis. An alert of abnormal bone in a chest X-ray report will inform the physician for early screening if appropriate and help prevent future fractures.
Chest X-rays are the most common imaging examinations done in healthcare settings, routinely used for investigating lung and heart diseases. Many of these X-rays are done in the emergency room and readily available for a PCP when the patient is seen for a follow-up visit. Plain radiographs are however, not very sensitive in detecting osteopenia or low bone mass as about 40% loss of bone mass is required to occur for osteopenia to be detectable on plain X-rays. 5 Signs of bone demineralization may be detected on X-rays usually as decreased cortical thickness and increased radiolucency as a result of loss of trabeculae. 6 Deep learning methods have been evaluated in detecting low BMD using plain radiographs.7,8 These radiographic features of bone loss are often overlooked or under-reported by many clinicians and radiologists on radiographs ordered for emergency and other non-skeletal causes.9,10 If routinely reported, plain radiographs can potentially help identify patients at risk of developing osteoporosis and fragility fractures. 11 Patients so identified could then undergo early screening using DXA scan, the gold standard screening test.
While the scientific literature discussed the possibility of identifying bone loss on plain radiographs, currently it is not standard practice for a radiology report to include demineralization status in the impression of the report for chest radiographs done for emergencies and other non-skeletal indications. Comments on demineralization are often overlooked in reports on these patients. Even when comments on bone demineralization are included, they are often buried in the body of the report and seldom make it to the final impression to draw the attention of the referring physician.9,12,13 Whether reporting demineralization on chest X-rays done for non-skeletal reasons would identify substantial number of cases that can potentially impact skeletal health outcomes in at risk patients has not been explored.
We aimed at comparing the prevalence of low bone density using chest radiographs obtained for any indication in patients 50 to 65 years old seen at an emergency department as assessed from their radiology reports and by an independent review by an emergency radiologist. We also sought to explore clinical and demographic predictors of low bone density among this population. This will enable us determine patients who may be at risk of osteoporosis and future fractures and would otherwise have been missed, as they fall outside the standard screening age.
Methods
Study Design
We conducted a cross-sectional study involving retrospective review of electronic medical records of patients 50 to 65 years, an age group that is not usually included in the routine osteoporosis screening guidelines. We identified all patients between 50 and 65 years, who had a chest X-ray obtained at the emergency department (ED) of a large Level 1 trauma center in an urban teaching hospital for any indication between May 1, 2021, and May 31, 2021. We extracted information on the indications for the X-rays. We categorized the indications into “Cardiopulmonary-related,” if it was for cardiovascular or respiratory system problem, “Falls and trauma,” if requested for investigation following a fall, trauma, or for suspected fracture, “Device placement,” if requested to define the positioning of intravenous lines or tubes, and “Other,” if X-rays were requested for other miscellaneous reasons such as epigastric pain, upper limb weakness, etc. We reviewed the X-ray reports for any mention of “bone demineralization, decreased bone density, osteopenia, or osteoporosis.” We also extracted information on demography (age, sex, race), clinical (hypertension, diabetes, use of corticosteroids), and behavioral characteristics (tobacco smoking and use of alcohol). Age was defined as the number of a patient’s completed years at the time the chest X-rays were obtained. Where a patient had more than one chest X-ray during the month of the study, the earliest records were selected. Race was categorized into White, Black, and “Other”. Race was described as “Other,” if a patient’s records show race/ethnicity as neither White nor Black. Race information were as stated in the medical records.
X-ray Assessment of Bone Density
A fellowship-trained emergency radiologist who was blinded to the radiology reports independently reviewed all the included chest X-rays to report on “demineralization status.” X-ray assessment of bone density was binary (either normal or bone demineralization). Bone demineralization was considered in patients with subjectively decreased osseous cortical thickness and/or increased trabecular bone radiolucency of the osseous structures included within the posterior-anterior (PA) X-ray field of view. 4
Data Analysis
We analyzed the data descriptively summarizing the categorical variables using frequencies and percentages and continuous variables using means and standard deviations. We calculated the percentage of patients who had comments on “bone demineralization” on the X-ray reports in the charts and after the review by the study radiologist.
Logistic regression analysis was conducted to explore determinants of low bone density in a fully adjusted model controlled for demographic, clinical, and behavioral characteristics. We analyzed the data separately for men and women.
Using U.S. census data, we also extrapolated the opportunity currently missed in identifying patients with low bone mass that can potentially be captured through introducing routine report of demineralization on chest X-rays done for the men and women between the ages of 50 and 64 years.
The study was approved by our Institutional Review Board and granted exemption from informed consent.
Results
Three hundred and ninety (390) eligible patient records were reviewed consisting of 201 (51.5%) males and 189 (48.5%) females. Mean age of patients was 58.2 (±4.6) years with a range of 50 to 65 years. A total of 128 (32.8%) were White, 104 (26.7%) were Black and all other races (Other) were 158 (40.5%). Most common indications for chest X-rays were cardiopulmonary issues (74.5%). Table 1 is a summary of the characteristics of the study population. Figure 1 shows the indications for the chest X-rays.
Characteristics of Study Population.
Indications for chest X-rays.
The review of the patient records revealed that 4 patients (all women) had bone demineralization comments in their X-ray reports. On independent review by the study radiologist specifically for signs of bone loss, 49 patients (12.6% of total study population), 6 males and 43 females were identified as having signs of bone loss. Of the 49 who were identified by the radiologist as having low bone density, 2 had a “bone demineralization” comment in their reports.
Bivariate analysis showed that in women, a history of hypertension and increasing age were associated with increased risk of low bone density, with odds ratio (OR) and 95% confidence interval (CI) of 2.56 and 1.10 to 5.95 and 1.08 and 1.00 to 1.17, respectively. However, in a multivariable regression model only hypertension remained significant, OR = 2.41, 95% CI = 1.03 to 5.64. No associations were found in men.
According to the 2020 US population census, the total population of the USA is 333,287,562 with 36% (120,785,304) in the group 50 years and older, and about 19.4% (64,131,920) in age group 50 to 64 years. 14 Bone Health and Osteoporosis Foundation (BHOF) documents that about 43.4 million of the US adult population 50 years and older are suffering from low bone mass, and are at risk of developing osteoporosis and fragility fractures. 15 There is however, no data on the number of adults suffering from low bone mass specifically in the age group matching our study population (i.e., 50-65 years old). From the radiologist’s review of the chest X-rays, 12.6% of our population had bone demineralization. Adjusting our prevalence of bone demineralization for sex using the 2020 US census data as the standard, the sex-adjusted prevalence as determined by the radiologist review was 13.1%. Applying our adjusted prevalence of 13.1% to the 64.1 million adults 50 to 64 years in the US population, we estimate that about 8.4 million people will have bone demineralization in this population group (see Figure 2). That means approximately 8.4 million people aged between 50 and 64 years may have undiagnosed bone demineralization and could potentially be identified as at risk for osteoporosis before they become eligible for screening.
Estimated number of cases of low bone mass among adults 50 years and older and the group between 50 and 64 years of age in the USA.
Discussion
Our study found that 49 of the total 390 (12.6%) of our population of adults 50 to 65 years showed signs of bone loss and could be at risk of osteoporosis. Eighty-eight percent (43/49) of these at-risk adults were women. These individuals fell outside the recommended age range for osteoporosis screening and would not have been identified using standard screening guidelines.
In this study, 47/49 (95.9%) of cases with low bone density were not identified or reported in the original X-ray reports. We also found that in women, a history of hypertension was associated with an elevated risk of low bone density (per radiologist review), although no such associations were found in men. Our study suggests that about 8.4 million adults between 50 and 65 years may be at risk of osteoporosis and subsequent fractures but may be undiagnosed because they remain outside the recommended age for osteoporosis screening. Although it may not be cost-effective to screen all adults 50 to 65 years for osteoporosis using the gold standard DXA scans, introducing a practice of reporting demineralization status in chest X-rays done for non-skeletal reasons may potentially identify some of the over 8 million people at risk for developing osteoporosis. These are patients who would otherwise be missed due to being outside the age (50-64 years) recommended in screening guidelines. Low screening rates with up to 80% oblivious of their risk of a fragility fracture, is a grave concern in addressing this chronic disease in older adults. 16
Opportunistic use of imaging modalities in screening for osteoporosis has been discussed extensively in the literature.11,17 Many studies have described its use for the diagnosis of osteoporotic vertebral fractures.11 -13,18 However, many of those studies identify osteoporosis after an osteoporotic fracture, usually a vertebral fracture had occurred. Our use of chest X-rays was to identify those at risk so that more definitive diagnostic processes could be initiated.
Our finding that 96% (47 out of 49) of patients with bone loss identified on independent review had no mention of bone loss in their X-ray reports is consistent with previous observations. A study in China looking at vertebral fractures using chest X-rays taken for other indications found that nearly 99% did not report vertebral fractures in the original X-ray reports. 9 Another study by Bazzocchi et al 19 to evaluate localization sequences of breast MRI as a potential tool to detect osteoporotic vertebral fractures, 71/799 (8.9%) patients were found to have vertebral fractures although 63 of the 71 (88.7%) had no previous history or diagnosis. Inattention blindness has been ascribed as a reason why experts may fail to detect abnormalities on images as they were not looking for these. 20 Another reason may be the phenomenon of “satisfaction of search” in which one is essentially satisfied at finding a lesion relevant to the search (indication based on the clinical order) and other lesions remain undetected.20,21 Another reason for under-reporting is that the majority (over 90%) of chest X-rays obtained through the ED on critically ill patients are “single, frontal view” radiographs and not the standard 2 views (frontal and lateral) thus limiting the opportunity to properly evaluate the thoracic bony cage and the thoracic spine. In this study, we think bone demineralization was under-reported on chest X-rays obtained for emergency and non-skeletal causes because the focus of the radiologists was more on the critical issues the patients were presenting with. This problem may be ameliorated by specifically requesting radiologists to pay attention to signs of bone loss particularly in women at elevated risk but may be missed in routine screening, for example, age group 50 to 64 years of age.
It is worth noting that 2 patients whose X-ray reports mentioned ‘demineralization’ were not identified as such by the study radiologist. This discrepancy may reflect the lack of standardized criteria for assessing low bone mass on X-rays, the inherent subjectivity of interpretation, and the non-specific nature of imaging findings. Nevertheless, opportunistic use of X-rays for pre-screening low bone mass in primary care may still offer clinical value.
Although hypertension is a known risk factor for osteoporosis, our statistical analysis failed to show significant association with bone demineralization in male participants.4,22,23 This might be secondary to the small number of males with bone demineralization in our cohort (i.e., 6 in males versus 43 in females). Women in our study group were from the peri- and early menopausal age groups in which bone loss is remarkably high due to fall in estrogen level. It is possible that hypertension which is postulated to increase bone loss may exacerbate bone loss in women at that age but not in men. It is outside of scope of this paper to investigate whether early onset of bone loss due to lack of estrogen in women made this group in our study more susceptible to bone loss in association with hypertension. It is also known that bone loss in men starts later in life than women; and this may have been the reason we saw such a small percentage of men having any sign of bone loss in their X-ray report as reviewed by the study radiologist.
It will be greatly beneficial for the patient if the radiologist includes a comment on bone mass on all the cases and include the finding of positive bone demineralization in the “impression” of the report to grab the attention of the referring clinician, especially in primary care settings, where the patients can at least be followed up on outpatient basis with DXA scans to identify cases of osteoporosis early and manage appropriately. Implementation would require training to raise radiologists’ awareness about commenting on bone structure in older adults, even when the X-ray is performed for reasons unrelated to osteoporosis. Given the high volume of chest X-rays conducted in healthcare settings, this could place a considerable burden on healthcare systems and providers. However, it may prove cost-effective by enabling earlier detection of osteoporosis and preventing related complications.
Our study had some limitations. This was a single center study, and the records reviewed were over 1 month. Chest X-ray is not an optimal imaging modality for identifying bone demineralization although it can alert the primary care physician. Assessing low bone mass due to age-related bone loss on X-rays is more “subjective” and interpretations can vary among radiologists especially on borderline and not so obvious cases. However, an opportunistic use of chest X-ray may be meaningful to identify patients in places where patients are not able to access a bone density screening easily and when a PCP suspects the risk of bone loss in a patient who is out of screening scope for a younger age. It is important to mention here that majority of the chest X-rays obtained through ED are single anteroposterior/upright views only, as positioning for lateral view on critically ill patients is challenging. Our data shows nearly 40% of the subjects had race as “Other” because we classified all races other than Caucasians and Blacks as “Other.” Although this will limit our ability to make generalizations concerning race, we do not believe it compromises the external validity of our findings, as we did not estimate race-specific measures.
Conclusion
Although mass screening for osteoporosis or bone demineralization in all adults 50 to 65 years is not the current recommendation, making use of plain chest radiographs ordered for other indications could provide a means of identifying patients at risk at little or no added cost to the patient. Looking for signs of bone loss particularly in women in their 50s and 60s may help improve rate of detection of bone loss on chest X-ray in this group who has additional risk factors, for example, lack of estrogen, hypertension, outside of simple aging process. Use of opportunistic chest X-ray may be a parsimonious yet pragmatic use of resources in identifying patients at risk of bone loss outside the traditional screening age.
Footnotes
Ethical Approval and Informed Consent
This study was approved by our Institutional Review Board (HSC-MS-22-0868 ) and informed consent was waived.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Data Availability Statement
Data used for this analysis is available from the corresponding author on reasonable request