Frequency of bone mineral density testing in adult kidney transplant recipients from Ontario, Canada: a population-based cohort study
We lack consensus on the clinical value, frequency, and timing of bone mineral density (BMD) testing in kidney transplant recipients. This study sought to determine practice patterns in BMD testing across kidney transplant centres in Ontario, Canada, and to compare the frequency of testing in kidney transplant recipients to non-transplant reference groups.
Using healthcare databases from Ontario, Canada we conducted a population-based cohort study of adult kidney transplant recipients who received a transplant from 1994-2009. We used logistic regression to determine if there was a statistically significant difference across transplant centres in the decision to perform at least one BMD test after transplantation, adjusting for covariates that may influence a physician’s decision to order a BMD test. We used the McNemar’s test to compare the number of recipients who had at least one BMD test to non-transplant reference groups (matching on age, sex, and date of cohort entry).
In the first 3 years after transplant, 4821 kidney transplant recipients underwent 4802 BMD tests (median 1 test per recipient, range 0 to 6 tests), costing $600,000 (2014 CAD equivalent dollars). Across the six centres, the proportion of recipients receiving at least one BMD test varied widely (ranging from 15.6 to 92.1 %; P < 0.001). Over half (58 %) of the recipients received at least one BMD test post-transplant, a value higher than two non-transplant reference groups (general population with a previous non-vertebral fracture [hip, forearm, proximal humerus], 13.8 %; general population with no previous non-vertebral fracture, 8.5 %; P value <0.001 for each of the comparisons).
There is substantial practice variability in BMD testing after transplant. New high-quality information is needed to inform the utility, optimal timing, and frequency of BMD testing in kidney transplant recipients.
KeywordsBone mineral density Kidney transplant recipient Bone
bone mineral density
chronic kidney disease-mineral and bone disorder
Institute for Clinical Evaluative Sciences
Kidney Disease Improving Global Outcomes
Ontario Health Insurance Plan
Mise en contexte
À ce jour, il n’existe aucun consensus sur la pertinence, au plan clinique, de demander une analyse de la densité minérale osseuse (DMO) chez les receveurs d’une greffe de rein, non plus que sur la fréquence ni le moment opportun pour y soumettre les patients après leur intervention.
Objectifs de l’étude
L’étude avait pour but d’établir un schéma de pratique pour la mesure de la DMO dans plusieurs centres de transplantation rénale en Ontario, au Canada. On a également voulu comparer la fréquence de ces analyses chez les patients ayant reçu une greffe rénale par rapport à un groupe de référence constitué de patients non transplantés.
Cadre et type d’étude
Il s’agit d’une étude rétrospective par cohorte représentative de la population, qui s’est tenue dans six centres de transplantation rénale en Ontario, au Canada.
Il s’agit d’une cohorte de patients ayant reçu une greffe du rein entre 1994 et 2009.
Les renseignements sur la fréquence, le coût total et les variations dans le nombre d’analyses de la DMO pour une période couvrant les trois années suivant la greffe ont été compilés dans chacun des six centres. La fréquence des analyses de la DMO chez les patients greffés a été comparée à la fréquence de ces mêmes tests pratiqués chez les sujets de groupes témoins, apparentés sur les plans de l’âge, du sexe et de la date de leur admission dans la cohorte, mais n’ayant pas subi une greffe du rein.
L’analyse par régression logistique a été utilisée pour établir la présence de différences significatives du point de vue statistique entre les six centres de transplantation en regard de la décision d’effectuer au moins un test de DMO à la suite d’une greffe rénale. L’analyse a tenu compte des covariables qui pouvaient influencer les médecins traitants au moment de décider de procéder ou non à un test de DMO sur leurs patients greffés. Le test McNemar a été utilisé pour comparer le nombre de patients greffés ayant été soumis à une analyse de leur DMO par rapport au groupe témoin.
À l’intérieur d’une période de trois ans suivant leur transplantation, un total de 4802 analyses de DMO ont été demandées parmi les 4821 patients greffés du rein répertoriés dans les six centres participant à l’étude. La valeur médiane se situait à un test par patient sur une échelle allant de 0 à 6 tests par patient. Le coût total évalué pour ces 4802 analyses de DMO était de 600 000 CDN en 2014. La proportion de receveurs de greffe ayant été soumis à une analyse de leur DMO a fluctué considérablement d’un centre de transplantation à l’autre, avec des pourcentages variant de 15,6 % à 92,1 % (P < 0,001). Dans l’ensemble, on a analysé la DMO de plus de la moitié (58 %) des patients greffés au moins une fois après leur intervention. Ce résultat s’est avéré plus élevé que les pourcentages mesurés dans deux des groupes témoins non transplantés (valeur de P < 0,001 pour chaque cas) : un premier groupe constitué de gens qui avaient subi une fracture non vertébrale (hanche, avant-bras ou humérus proximal) par le passé (13,8 %) et un second groupe constitué de gens de la population générale n’ayant pas subi de fractures (8,5 %).
Limites de l’étude
Les renseignements concernant les médicaments d’ordonnance administrés aux participants étaient incomplets et les valeurs de DMO étaient manquantes dans plusieurs cas. De plus, le faible taux de fractures subies par les participants ne permet pas d’établir une relation entre la valeur de DMO mesurée et le risque de fractures.
Une variabilité importante a été constatée dans la pratique d’analyses de la DMO à la suite d’une transplantation rénale. Davantage de données sont nécessaires pour discuter de la pertinence d’effectuer ce test chez les receveurs de greffe rénale, ainsi que pour établir le moment opportun et la fréquence à laquelle les y soumettre après l’intervention.
What was known before
Kidney transplant recipients have a higher risk of fracture compared to the healthy general population. However, the best way to identify recipients at high risk is unknown. Bone mineral density (BMD) is widely used in the general population to help identify patients with a high fracture risk, but its utility in the kidney transplant population is unclear. Limited evidence can lead to substantial practice variability. However, practice patterns for BMD testing in kidney transplant recipients from Ontario, Canada, are unknown.
What this adds
Despite limited evidence on the utility of BMD testing, many kidney transplant recipients receive a test in the 3 years after transplant, and significant practice variability in BMD testing was observed across transplant centres. These results indicate future studies are needed to better understand the utility, frequency, timing, and cost-effectiveness of BMD testing in kidney transplant recipients.
It is well accepted that kidney transplant recipients have a higher risk of fracture compared to the healthy general population [1, 2, 3, 4]. For example, Ramsey-Goldman et al., found female kidney transplant recipients aged 25–44 years had an 18 times higher fracture risk compared to individuals from the general population of similar age and sex ; however, recent observations suggest that the absolute incidence is low with only 1.7 % of recipients sustaining a hip fracture in the 10-years after transplantation (high risk defined as a 10-year risk ≥3 %) [4, 5, 6, 7]. The reasons for this higher risk are multifactorial and include pre-existing chronic kidney disease-mineral and bone disorder (CKD-MBD) and glucocorticoid administration after transplantation . In the general population, Osteoporosis Canada guidelines recommend bone mineral density (BMD) testing is done in individuals at a high risk of fracture, as a decreased BMD can help risk stratify those individuals at higher risk of fracture [9, 10, 11]. However, in the kidney transplant population, the ability of BMD to predict fracture is unclear [12, 13, 14]. Limited evidence can lead to substantial practice variability. Therefore, we conducted a population-based study to determine the frequency, total cost, and the variability in BMD testing across all six transplant centres in Ontario, Canada. To help put the frequency of BMD testing into context, we also compared the frequency of testing in transplant recipients to non-transplant reference groups (matching on age, sex, and date of cohort entry).
Design and setting
We used linked healthcare databases from the province of Ontario, Canada to conduct this study. Universal access to physician and hospital services is provided to all Ontario residents. These datasets were linked using unique encoded identifiers and analyzed at the Institute for Clinical Evaluative Sciences (ICES). This study was approved by the institutional review board at Sunnybrook Health Sciences Centre, Toronto, Canada.
Information on Ontario kidney transplant recipients is provided by the Canadian Organ Replacement Register. Information on Ontario physicians’ billing claims for inpatient and outpatient services is reported by the Ontario Health Insurance Plan (OHIP). The Ontario Registered Persons Database provides information on demographics and vital status. Prescription drug utilization data is provided from the Ontario Drug Benefit Plan, which is a universal drug plan for individuals ≥65 years. It also provides information since April 1997 on special populations <65 years who are eligible for the program. The ICES Physician Database provides information on physician specialty. Emigration from the province was the only reason for loss to follow-up (0.5 % per year) .
Kidney transplant recipients
We included all adults (age ≥ 18 years) with a first kidney transplant from July 1, 1994 to December 31, 2009. We defined the date of the kidney transplant as the date of cohort entry (also referred to as the index date).
To put the frequency of BMD testing into context, we matched recipients on age (±1 year), sex, and index date (±1 year) to two non-transplant reference cohorts; one group was considered to be at a low fracture risk where we would not expect frequent testing and one group was at an increased fracture risk where we would expect more frequent testing. Specifically, we matched recipients to the general population with no previous non-vertebral fracture (low fracture risk) (defined as proximal humerus, forearm, hip) and the general population with a previous non-vertebral fracture (increased fracture risk). When permitted by the available sample, we matched one recipient to four persons from the non-transplant reference cohort. Further detail on the cohort creation for these reference groups has been described elsewhere . In Ontario, the 3-year incidence of fracture is higher in individuals with a previous non-vertebral fracture than in kidney transplant recipients (of which 1.6 % have a non-vertebral fracture [proximal humerus, forearm, hip] post-transplant) .
We used physician fee-for-service billings to identify BMD by dual energy X-ray absorptiometry and, prior to April 1998, dual-photon absorptiometry tests (Additional file 1) . In Ontario, these data are largely complete with approximately 94 % of physicians submitting such billing . These BMD billing codes have been successfully used in several prior studies [18, 19]. We tabulated the number of BMD tests in the 3 years following kidney transplantation; multiple billings for a BMD test for a given person on the same day were counted as one test. We selected a 3-year follow-up to allow enough time for recipients to undergo multiple BMD tests; Osteoporosis Canada guidelines recommend a repeat BMD test in the 1–3 years after the initial test . To calculate the total cost of the BMD tests, we included all associated billings (technical component of the test and professional component [e.g. physician interpreting the BMD test]) and accounted for inflation; additional information on billings can be found in Additional file 2.
We used medians (interquartile range [IQR]) or means (standard deviation) to summarize baseline characteristics for continuous data and proportions to summarize categorical data. To compare baseline characteristics between recipients with at least one BMD test to those without a BMD test, we used the chi-square, Mann-Whitney U test, or Student’s t test as appropriate. We stratified the frequency of BMD testing by sex (men versus women) and age at the time of transplantation (<50 versus ≥50 years). We selected this age dichotomization for several reasons: kidney transplant recipients aged ≥50 years are at an increased fracture risk compared to younger recipients [4, 20]; favourable statistical properties (median age of our cohort was 50 years); and Osteoporosis Canada guidelines recommend BMD testing in individuals aged 50–64 years who have clinical risk factors for fracture (e.g. prolonged high-dose glucocorticoid use) . We used logistic regression to determine if there was a statistically significant difference across transplant centres in the decision to perform at least one BMD test after transplantation. We adjusted for covariates that may influence a physician’s decision to order a BMD test (age, sex, previous non-vertebral fracture, and comorbidities [as measured by the Charlson comorbidity index ]). To determine if there were changes over time in the number of BMD tests performed, we used the Cochran-Armitage test for trend. To compare the number of recipients who had at least one BMD test to the matched non-transplant reference groups, we used McNemar’s test. We considered a two-sided p value <0.05 as statistically significant. We performed all analyses using the Statistical Analysis Software (SAS version 9.3).
Baseline characteristics of kidney transplant recipients classified by presence of at least one bone mineral density test in the 3 years after transplantationa
Bone mineral density test
(n = 2786)
(n = 2035)
1182 (42 .4 %)
599 (29.4 %)
290 (10.4 %)
624 (30.7 %)
631 (22.6 %)
480 (23.6 %)
769 (27.6 %)
413 (20.3 %)
1096 (39.3 %)
518 (25.5 %)
690 (24.8 %)
565 (27.8 %)
Previous non-vertebral fracture
68 (2.4 %)
41 (2.0 %)
Charlson comorbidity indexb
2.6 ± 1.0
2.7 ± 1.2
Bone mineral density
Number (proportion) of kidney transplant recipients with at least one bone mineral density test in the 3 years after transplantation by age and sex
Kidney transplant recipients
(n = 4821)
2786 (57.8 %)
Women <50 years
612 (64.8 %)
Women ≥50 years
570 (68.1 %)
Men <50 years
741 (50.7 %)
Men ≥50 years
863 (54.7 %)
Non-transplant reference groups
Number (proportion) of kidney transplant recipients with at least one bone mineral density test in the 3 years of follow-up compared to several reference groups matched on age, sex, and index date
Kidney transplant recipients (n = 4821)
2786 (57.8 %)
General population with no previous non-vertebral fracture (n = 19,284)
1645 (8.5 %)
General population with a previous non-vertebral fracture (n = 4821)
665 (13.8 %)
Of the 3540 recipients who had prescription drug coverage through universal healthcare benefits, 646 (18.2 %) were prescribed bisphosphonates. Of recipients prescribed bisphosphonates, 548 (84.8 %) of these prescriptions were filled at a median of 57 days (IQR 21 to 175 days) after the BMD test, with 417 receiving a bisphosphonate prescription in the first 6 months after a BMD test.
In Ontario, Canada, we found that over half of the kidney transplant recipients underwent at least one BMD test in the subsequent 3 years after transplant, and many recipients underwent multiple tests. The frequency of BMD testing varied widely by centre—from as few as 15 % of recipients undergoing a BMD test to as many as 92 %, and this variability was not explained by recipient characteristics. Kidney transplant recipients were significantly more likely to undergo a BMD compared to two matched non-transplant reference groups. Our results suggest that BMD testing is commonly performed in kidney transplant recipients despite weak evidence in the literature supporting its widespread use.
The results of our population-based multicentre study extend the findings of two prior single centre reports with smaller sample sizes. In the first study of kidney transplant recipients (n = 326) from Manitoba, Canada, almost 60 % of recipients were found to have had at least two BMD tests within approximately 8 years of their transplant . The second study from Akaberi et al. found that 670 BMD tests were performed in 238 kidney transplant recipients (75 % had at least two BMD tests) from Sweden over 12 years . The centres in these two prior studies had protocols in place for routine BMD testing, and so the frequency of BMD testing would be expected to be high. In contrast, in our study, only a few of the transplant programs had a protocol for BMD testing (information provided by the six Ontario transplant centres, personal communication).
Particularly striking is the high number of kidney transplant recipients who underwent multiple BMD tests in the 3 years after transplantation, at a high cost to the healthcare system. For example, almost one third of kidney transplant recipients underwent two or more BMD tests within 3 years of their transplant; in the non-transplant population, the benefits of performing multiple BMD tests over several years have been questioned [23, 24], especially given the increasing knowledge of unwarranted screening harms [25, 26].
The benefit of BMD tests in kidney transplant recipients remains uncertain. First, the utility of BMD in predicting fracture in kidney transplant recipients is unclear [12, 13, 14]. For example, the Kidney Disease Improving Global Outcomes (KDIGO) guidelines for CKD-MBD suggest that patients with an estimated glomerular filtration rate >30 mL/min/1.73 m2 have their BMD assessed in the first 3 months after kidney transplant if they received glucocorticoids or have other risk factors for osteoporosis . However, given the limited evidence, this suggestion was given the weakest grade of evidence . It is important to note that this recommendation is currently being reassessed in the revised version of the guidelines  in light of recent evidence finding that BMD may be predictive of fracture in individuals with CKD, including dialysis [28, 29]; however, there is still conflicting evidence in kidney transplant recipients [12, 13, 14]. Second, given the high incidence of adynamic bone disease (i.e. low turnover) in kidney transplant recipients, the KDIGO guidelines suggest that a bone biopsy may be needed to guide treatment decisions; this limits the clinical usefulness of BMD testing post-transplant . Third, and perhaps most relevant, recent research suggests in contrast to what has been previously reported, most kidney transplant recipients will not fracture and have an average mean BMD for age and sex [4, 5, 12, 22, 30]. Note, however, that the lower than expected fracture incidence and normal BMD may be the result of increased monitoring of bone health after transplant. Taken together, this suggests there may be little need to perform BMD tests routinely.
It is important to note that BMD testing may alter clinical practice. Many transplant recipients were prescribed a bisphosphonate in the first 6 months after receiving a BMD test. However, the efficacy of this and other fracture prevention strategies in kidney transplant recipients remains uncertain .
We make several recommendations based on the findings in this study. First, given the uncertainty that exists in the ability of BMD to predict fracture in kidney transplant recipients, new measures that have been found to predict fracture independent of BMD should be examined. For example, the trabecular bone score assesses bone quality (microarchitecture) and has been found to predict fracture in the general population [32, 33, 34]. This measure could be useful at predicting fracture in kidney transplant recipients given bone quality is particularly affected in recipients with CKD-MBD . Second, new high-quality information from prospective observational studies and clinical trials is needed to guide the optimal recommended timing and frequency of BMD testing. Such studies should also assess the ability of BMD to predict fracture and its cost-effectiveness.
Our study has some limitations. First, we did not have drug dispensing information for the entire transplant cohort (only those who were covered by provincial drug benefits). While we were unable to characterize immunosuppression use at the patient level, during the time frame of this study, steroids were nearly universally prescribed at the Ontario transplant centres. Second, we only knew if a BMD was done, without information on the BMD value. However, the former supported the primary objective of this study—to determine the frequency of BMD testing in the first 3 years after transplant across several kidney transplant centres. Third, due to the low number of fracture events, we were not able to determine if transplant centres with more BMD tests had fewer fracture events. Moreover, the small number of recipients with a previous non-vertebral fracture may have limited our statistical power to determine whether recipients with a previous fracture were more likely to undergo a BMD test. Fourth, the variability in BMD testing we observed across transplant centres was in the setting of universal healthcare benefits. It is possible that these results may not generalize to other types of healthcare systems; variability across transplant centres might be even greater in jurisdictions without such healthcare benefits, as economic factors may also influence testing. Finally, we did not assess the impact of the KDIGO CKD-MBD guidelines on BMD testing. However, this guideline received the weakest grade of evidence; therefore, its uptake would likely be variable across transplant centres as demonstrated in this study.
Many kidney transplant recipients underwent a BMD test in the 3 years after transplantation despite the lack of evidence to suggest BMD can accurately predict fracture. These results highlight the need for further studies to understand the utility, frequency, timing, and cost-effectiveness of BMD testing in kidney transplant recipients.
Dr. Naylor is supported by the Canadian Institutes of Health Research Fellowship and the Canadian National Transplant Research Program Astellas Training Award. Dr. Lam is supported by a KRESCENT new investigator award. Dr. Garg was supported by the Dr. Adam Linton Chair in Kidney Health Analytics. This study was supported by the Institute for Clinical Evaluative Sciences (ICES), which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (MOHLTC). The opinions, results, and conclusions reported in this paper are those of the authors and are independent from the funding sources. No endorsement by ICES or the Ontario MOHLTC is intended or should be inferred. Parts of this material are based on data and information compiled and provided by CIHI. However, the analyses, conclusions, opinions, and statements expressed herein are those of the author, and not necessarily those of CIHI.
- 7.National Osteoporosis Foundation. Clinicians guide to prevention and treatment of osteoporosis. Washington, DC: National Osteoporosis Foundation; 2010.Google Scholar
- 8.KDIGO Work Group. KDIGO clinical practice guidelines for the prevention, diagnosis, evaluation, and treatment of Chronic Kidney Disease-Mineral and Bone Disorder (CKD-MBD). Kidney Int. 2009;S113:S1–S130.Google Scholar
- 15.Ontario Ministry of Finance. Ontario population projections update 2012–2036. Toronto: Government of Ontario; 2013.Google Scholar
- 16.Ontario Ministry of Health and Long term Care. Schedule of benefits. Ontario: Queen’s Printer; 2014.Google Scholar
- 17.Institute for Clinical Evaluative Sciences. Improving health care data in Ontario. Toronto: ICES; 2005.Google Scholar
- 27.Ketteler M, Elder GJ, Evenepoel P, Ix JH, Jamal SA, Lafage-Proust MH, et al. Revisiting KDIGO clinical practice guideline on chronic kidney disease-mineral and bone disorder: a commentary from a Kidney Disease: Improving Global Outcomes controversies conference. Kidney Int. 2015;87:502–28.PubMedCrossRefGoogle Scholar
- 31.Palmer SC, McGregor DO, Strippoli GFM. Interventions for preventing bone disease in kidney transplant recipients. Cochrane Database Syst Rev. 2009;4:1.Google Scholar
- 33.Rabier B, Heraud A, Grand-Lenoir C, Winzenrieth R, Hans D. A multicentre, retrospective case-control study assessing the role of trabecular bone score (TBS) in menopausal Caucasian women with low areal bone mineral density (BMDa): analysing the odds of vertebral fracture. Bone. 2010;46:176–81.PubMedCrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.