Three-phase ^99mTc-labeled bone scintigraphy is widely used for the evaluation of suspected inflammatory osseous conditions and specific clinical scenarios, including osteomyelitis and septic arthritis [1–5], inflammatory arthritis [6–9], prosthesis complications [8–10], diabetic foot [11–13], osteoid osteoma [14], complex regional pain syndrome [15, 16], and sports-related injuries [17]. Three-phase bone scanning involves targeted imaging of the area of interest in the angiographic (flow, immediately after injection), soft-tissue (pool, 2–10 minutes after injection), and delayed (2–4 hours after injection) phases. At our institution, the entire skeleton in addition to the area of interest is imaged during the delayed phase. The literature contains little evidence supporting this practice [18], but the findings provide additional information without increasing the radiation dose to the patient. In addition, the billing code and therefore direct financial cost to the patient associated with limited three-phase bone scintigraphy do not change when whole-body imaging is added to the examination. Delayed whole-body imaging may therefore improve overall diagnostic utility without imparting substantial cost to the patient.

Despite the apparent risk- and expense-free benefits, imaging the entire skeleton in the delayed phase consumes additional time, effort, and institutional resources. Bone scintigraphy is well known to be a sensitive but nonspecific examination [19–23]. A lesion incidentally detected in an asymptomatic focus (low pretest probability of disease) may be unrelated to the clinical question. In many cases such a lesion prompts further clinical and diagnostic evaluation that consumes temporal and financial resources [24]. In addition, imaging the entire skeleton increases gamma camera time, which can decrease institutional efficiency. Likewise, there is a concomitant increase in interpretation time spent by radiologists. We hypothesize that for most indications, delayed whole-body imaging after directed three-phase bone scintigraphy does not improve diagnostic yield or alter patient care.

Before the initiation of this investigation, institutional review board approval was obtained. The study was performed in compliance with HIPAA. Informed consent from patients was not required owing to institutional policy and the retrospective nature of this investigation.

The study sample included 400 patients (156 men and boys, 244 women and girls; mean age, 45 years; range, 1–90 years). Fifty-two of the patients (13%) were 0–17 years old, and 236 (59%) were older than 40 years. The mean increase in examination duration due to use of whole-body imaging was 25 minutes (range, 21–31 minutes).

Table 1 lists the frequency of each category of clinical indications. Most of the requisitions (83%, 334/400) indicated only one or two foci of interest and were categorized as simple. Review of the medical records indicated that 13% (51/400) of the patients had a medical history of malignant neoplasm (Table 2). No pediatric requisitions identified a concern about osteomyelitis or septic arthritis. Tracer distribution was normal in 131 examinations (33%). Mild or moderate degenerative changes were found in 194 of the examinations (49%), severe degenerative changes in seven (2%), solely degenerative changes in 103 (26%), and findings unrelated to the area of interest in 166 examinations (41%). The mean age of the patients with severe degenerative changes was 73 years (range, 61–87 years).

The findings outside the area of interest did not alter diagnosis or diagnostic certainty about the three-phase findings in any case but did generate 82 recommendations for additional evaluation. Thirty-seven soft (9%) and 45 hard (11%) recommendations led to performance of the following (Table 3): 41 radiographic examinations, eight unspecified imaging studies, four CT examinations, 11 MRI examinations, two nuclear medicine examinations, two ultrasound examinations, one laboratory test, and 34 clinical correlations. Some recommendations included multiple courses of action. As a direct result of these recommendations, clinicians requested 18 radiographic, two CT, one MRI, one ultrasound, and one bone scintigraphic follow-up examinations and two referrals to a consultant (Table 3). Recommendations based on findings outside the three-phase area of interest affected treatment in only one case: Temporomandibular joint uptake resulted in a referral for physical therapy.

Six radiologists made final reports of the 400 three-phase bone scintigraphic examinations in this study. The ratio of recommendations made to number of examinations interpreted was as follows: radiologist 1, 37/145 (26%); radiologist 2, 16/77 (21%); radiologist 3, 10/56 (18%); radiologist 4, 10/61 (16%); radiologist 5, 5/44 (11%); radiologist 6, 4/17 (24%).

Three-phase ^99mTc-labeled methylene diphosphonate bone scanning is popular for the evaluation of certain pathologic conditions and clinical scenarios. It is routine practice at our institution to perform a delayed whole-body examination as part of a focused three-phase study. It can be argued that this practice provides free additional information because no additional radiation dose is used to examine the rest of the body. In a 2002 study, Yang et al. [25], using whole-body imaging performed in all three phases, detected 394 abnormalities during the angiographic and soft-tissues phases in 542 patients. They argued that whole-body three-phase imaging has utility in identifying numerous abnormalities that would not otherwise be detected with delayed whole-body imaging. Those investigators, however, did not specify the number of previously unknown abnormalities or the number of cases in which treatment was altered as a result of lesion detection. They did illustrate cases in which the three-phase data were useful in characterizing a lesion, but the findings simply indicate the usefulness of three-phase imaging irrespective of the field of view. In our series, the whole-body data contributed to only one instance of altered management in 400 consecutively evaluated cases. Furthermore, the discovery of 166 clinically irrelevant findings spawned 82 recommendations and an additional 25 expensive studies and referrals. The institutional and patient expense inherent in these evaluations should be considered in the assessment of the diagnostic appropriateness and overall outcome of delayed whole-body scintigraphy.

The mean increase in duration of bone scans due to whole-body imaging was 25 minutes. During the period of the study, this scanning contributed an additional 167 hours of gamma camera time and technologist work. We did not assess the time radiologists spent interpreting whole-body images, correlating the images with previous images, and contacting referring physicians to alert them to the findings because it was difficult to separate this resource utilization from that for a three-phase study alone. If the time lost obtaining the additional images had been recapitalized, it would have been possible to perform approximately 167 additional bone scans during the dates of the study.

After 20% (82/400) of bone scans in this study, a recommendation was proffered for additional evaluation of whole-body findings outside the three-phase area of interest. As a direct result of these recommendations, clinicians requested 23 additional diagnostic studies and made two referrals to consultants. In many cases, clinicians either did not take a course of action or took a course of diagnostic action different from that recommended by the nuclear radiologist. We did not document repeated patient visits to the ordering physician for discussion of the incidental findings because it was difficult to establish whether the scan findings were the primary reason for the visit. In all cases in which recommendations were made, the medical record was reviewed to determine the outcome of the findings in question. Lesion relevance was followed to the last clinical note or imaging study available. In only one case was an alternative or enhanced diagnosis made on the basis of the incidental findings (temporomandibular joint disorder). Likewise, treatment was altered only once, for this same case of temporomandibular joint disorder.

Findings detected outside the three-phase area of interest on the whole-body images did not alter diagnostic certainty or the primary scintigraphic diagnosis in any case. In addition, even though incidental findings outside the area of interest were seen in 41% (166/400) of examinations, none of the 82 recommendations regarding these findings affected morbidity or mortality. These data suggest that routine acquisition of delayed whole-body images after directed three-phase examinations is not indicated and is an unnecessary expenditure of health care resources. Interestingly, even in clinical scenarios in which one might postulate a high pretest probability of the presence of multifocal disease, the delayed whole-body findings were not useful. Thirteen percent (51/400) of patients who underwent imaging in this study had a history of malignant neoplasia; 59% (236/400) of the patients were older than 40 years; and 11% (44/400) of requisitions indicated the presence of multifocal symptoms.

There were several limitations to this study. Although 13% (51/400) of patients had a history of malignant neoplasia, the disease status and tumor burden at the time of the examination were not ascertained. In addition, the list of neoplasms was a heterogeneous set, as indicated in Table 2. Despite these limitations, we suggest that delayed-phase whole-body bone scintigraphy may already be part of the surveillance protocol for neoplasms with a high yield from bone scintigraphy, and obtaining additional images at unscheduled times to evaluate asymptomatic foci may not benefit the treatment of these patients. We cannot comment on the validity of combined whole-body and three-phase examinations for pediatric osteomyelitis because none of the pediatric patients in our study had that indication. This finding was surprising and may indicate either an anomaly in our referral pattern or that other imaging studies (e.g., MRI) are being used in lieu of bone scintigraphy. Pediatric osteomyelitis is well known to be hematogenously disseminated [26], and it is possible that imaging the entire skeleton may be of benefit to these patients. Because our study included only patients from our institution, referral bias may have affected our results. Last, it is possible that although rare, some lesions incidentally detected at whole-body imaging during three-phase examinations may be clinically significant and that missing these lesions can delay diagnosis and increase cost. It may be necessary to increase our sample size to identify these cases and determine what if any savings are imparted by their detection.

Our data are evidence that routine wholebody imaging after directed three-phase diphosphonate bone scintigraphy does not improve diagnostic yield, does not alter patient care, and is an unnecessary use of medical resources. Adding whole-body images may be beneficial in selected individual circumstances, but those circumstances should be the exception and not the rule. In this era of expense reduction and efficiency, radiologists should be mindful of the hidden costs of imaging and make an effort to maximize the yield of the studies they perform.

Address correspondence to M. S. Davenport ([email protected]).