Developing a Toolkit for Pediatric CT Dose Reduction in Community Hospitals

Abstract

“Eighty percent of success is just showing up.”

Woody Allen

Pediatric CT radiation dose optimization is a challenging process for pediatric-focused facilities and community hospitals alike. Ongoing experience and trial-and-error approaches to dose reduction in the large academic hospital setting may position these centers to help community hospitals who strive for CT quality improvement. In this pilot project, we describe our hands-on approach to develop a partnership between an academic medical center and a community hospital to develop a toolkit for implementing CT dose reduction. Our aims were to 1) assess the acceptability of an interactive educational program and electronic toolkit booklet, 2) conduct a limited test of the efficacy of the toolkit in promoting knowledge and readiness to change, and 3) assess the acceptability and practicality of a collaborative approach to implementing dose-reduction protocols in community hospitals. In partnering with the community hospital, we found that they had size specific radiation doses 2 to 3 times higher than those at our center. Survey results following a site visit with interactive educational presentations revealed an increase in knowledge, stronger opinions about the health risks of radiation from CT scans, and willingness and perceived ability to reduce pediatric CT doses.

Introduction

U.S. population doses from medical radiation sources are at a historic high, having increased by 600% since the 1980s [1-2]. The greatest contributor to the striking increase in population exposure is the CT scan [3]. In recent years, CT scans have drawn considerable scrutiny due to the health risks of ionizing radiation [3-7]. Two epidemiologic analyses in the U.K. and Australia on large populations over many years have shown similar results: a 24% increase in cancer in children and adolescents exposed to ionizing radiation from CT scans [8-10]. These lifetime risk estimates based on direct analyses and epidemiologic data gleaned from national health registries are compatible with various lifetime risk estimates derived from atomic bomb survivor data [4, 7]

Children are especially vulnerable to the harmful effects of radiation. For example, risk projections suggest that for an abdominal or pelvic CT scan, the lifetime risks for children are 1 cancer per 500 scans, regardless of age at exposure [11, 12]. Radiation-induced cancer risk is magnified in children and adolescents because there is more time for the malignancy to manifest (up to decades later) and because their cells divide more rapidly, making them more susceptible [7, 13]. A 2013 study in JAMA Pediatrics [14] culled data from a large research network of health maintenance organizations in the United States and found that many children received high radiation doses from CT scans. The authors attribute this finding both to the greater use of higher-dose CT examinations, such as scans of the abdomen and pelvis, and to substantial variability in radiation doses [2, 14]. For example, in one study, among children enrolled in a health maintenance organization, the doses for abdominal CT in children 1–4 years old was 4.8–137 mSv for a single CT scan [13, 15]. They project that if radiation doses nationwide reflect the doses they observed for CT scans of the head, abdomen/pelvis, chest, and spine for children younger than 15 years, then the scans performed in 1 year in the United States might cause 4,870 future cancers [14]. The authors suggest that if the highest 25% of doses can be reduced to the median dose, then 43% of those cancers might be prevented [14].

The practice of using the same radiation exposure factors for CT examinations of children as those for adults is not uncommon [16, 17], and continues today, particularly in non-pediatric focused facilities [18]. Of the estimated 8 million CT studies performed on children, up to 6.8 million are performed at sites outside the auspices of a dedicated pediatric facility [19]. Since 2008, the volume of CTs performed at children's hospitals and pediatric emergency departments has reached a plateau and presently may be decreasing, credited in large part to dose reduction technology in new CT scanners and educational strategies and social media campaigns promoted by professional societies such as the American College of Radiology (ACR) Imaging 3.0 Program and the Alliance for Radiation Safety in Pediatric Imaging [18-21]. Notably, the radiation burden of these exams has also decreased [19, 22]. However, the volume of CTs performed at non-pediatric focused facilities has increased [13, 19].

For the 6.8 million CT studies performed outside the auspices of a dedicated pediatric facility, there are critical questions about whether diagnostic reference levels (DRLs) are followed for radiation doses on pediatric CTs [19]. For example, a recent study at Children's Hospital Boston and Harvard Medical School evaluated abdominal-pelvic CT performed in children during trauma evaluations at community hospitals (CHs) prior to transfer of care [23]. The study revealed that fully one-half of children received radiation doses that were greater than the 75th percentile norm, ranging overall from 0.17 to 5.07 times the norm [23]. As initially described in 2001 by Paterson et al. [24] in pediatric patients undergoing abdominal CT, multiphase scanning is still common in the CH setting, despite the fact that, rather than increasing the diagnostic yield, radiation dose is substantially increased (up to tripled) [19, 25].

The purpose of this study was to develop a partnership between a large academic medical center (AMC) and a CH to develop a toolkit for implementing CT dose reduction, which eventually could be disseminated to CHs statewide. Our aims were to 1) assess the acceptability of an interactive educational program and electronic toolkit booklet, 2) conduct a limited test of the efficacy of the toolkit in promoting knowledge and readiness to change, and 3) assess the acceptability and practicality of a collaborative approach to implementing dose-reduction protocols in CHs.

Methods

Preliminary data

We conducted our research at a public, tertiary care academic, Level I, adult and pediatric trauma center. For the purpose of this study, a CH was defined as a general non-university hospital that does not specifically focus on the care of pediatric patients [26]. The biomedical institutional review board at our institution determined this study to be a category 4 exemption under 45 CFR 46.101(b).

In order to establish baseline data, our study team conducted a preliminary retrospective analysis using CT dose length product (DLP) contained in head CT imaging dose reports from 20 CHs and radiology practices sent to our center for second opinions or transfer of care during a 6-month period in 2012. This survey identified 12 sites with a variety of CT scanners and pediatric exposures (based on CT DLP) higher than our usual adult doses, with many sites using 2-10 times higher dose than our corresponding age-based protocols (see Data Supplement S1 for graphical representation of data).

Field Test

After identifying the variations in doses, the research team employed survey and semi-structured interview methods, to develop, assess, and refine a pediatric CT dose reduction toolkit in one CH that could be generalized over different manufacturers and platforms. Our research team made site visits to the CH and met with selected stakeholders, including the CH imaging administrative director, CT technologists, and private practice radiologists. In these meetings, we presented a brief, interactive educational program, discussed the project, and assessed knowledge of radiation-exposure and radiation-related cancer risk through pretest and post-test surveys (see Data Supplement S2 for survey/questionnaire). Pre- and post-test surveys were designed to explore attitudes, including opinions about CT risk, prior to and following the educational intervention on a 5-point Likert scale. These responses were totaled and compared across respondents in a pre/post-test fashion. The survey assessed staff familiarity with best practice guidelines in pediatric CT dose reduction, comfort level with the ALARA principle and preservation of diagnostic image quality, knowledge of automatic dose-reduction applications in existing CT systems, and familiarity with dose-reduction techniques on the Image Gently website.

We then distributed a pediatric dose reduction electronic toolkit booklet, The ABCs of Childcare in CT: Awareness, Belief, Change to CH participants. The electronic toolkit booklet included selected examples of pediatric protocols; selected medical literature regarding practical strategies for dose optimization with links to online articles; examples of head and abdominal CT with as low as reasonably achievable (ALARA) doses (see Data Supplement 3); glossary of definitions and terms, selected educational links, (for example Image Gently [27], Image Wisely [28] the ACR National Radiology Data Registry [29]; tips, contacts, links provided by the ACR Imaging 3.0; and a list of contact information for help from the AMC team.

We conducted semi-structured interviews with the radiology team to gather feedback about the current content of the educational intervention and toolkit as well as to identify issues related to dose optimization that may not be addressed by the educational session and toolkit. The interviews were transcribed and then reviewed for insights that supplement our survey data (see Data Supplement S4 for semi-structured interview format).

To establish baseline data for a future dose reduction intervention, we collected and quantified recent prior dose data from CH pediatric CT scans for comparison to benchmarks. Shared CH data included CT scanner vendor and scanning parameters obtained from data in the DICOM dose reports (see Data Supplement S5 for CT dose template).

Results

Comparative Baseline Doses CH versus AMC

CH radiation doses were approximately 2 to 3 times that of matched patients at AMC. The greatest variation occurred in the Head CT 0-3 age group, with a mean CTDI of 14 (95% CI: 11-17) and 54 (95% CI: 20-88) at AMC and CH, respectively. (Table 1)

Table 1.

Tabular comparison of mean pediatric size specific dose estimate (SSDE) and CTDI between Academic Medical Center (AMC) and Community Hospital (CH).

Abdomen Data
Age Category	Facility	SSDE	p Value	Significant
0 - 10	AMC	3.9 (3.0 - 4.8)	<0.001	Yes
	CH	9.6 (7.3 - 11.9)
10-17	AMC	9.7 (7.3 - 12.0)	<0.001	Yes
	CH	21.3 (15.8 - 26.8)
Head Data
Age Category	Facility	Mean CTDI	p Value	Significant
0 - 3	AMC	14.0 (11.0 - 17.0)	0.002	Yes
	CH	54.3 (20.0 - 88.5)
3 - 6	AMC	13.0 (7.6 - 18.4)	0.003	Yes
	CH	44.3 (27.0 - 61.6)
7 - 10	AMC	16.7 (11.5 - 21.8)	0.401	No
	CH	21.4 (15.2 - 27.6)
10 - 17	AMC	22.5 (16.0 - 29.0)	0.001	Yes
	CH	44.3 (31.4 - 57.1)

Interactive Educational Presentation Pretest/Post-test Survey

Results showed a small increase in opinions towards the importance of risks associated CTs as shown in Table 2 (mean difference 0.5 on 5 point likert scale). Given the small sample size, tests of statistical significance were not conducted. The next set of questions assessed knowledge of CT risks. Based on the responses to these 5 questions, we saw an increase in the knowledge related to CT risks for all participants after our intervention (mean difference 47%). All respondents were able to successfully answer more questions correctly following the educational intervention (see Table 2 and Data Supplement S2). In addition, we assessed the readiness for practice change (i.e., perceived change commitment and change efficacy) using items from the Organizational Readiness for Implementing Change (ORIC) instrument [30]. Results suggested that participants perceived both change commitment (mean 5 on 5 point likert scale) and change efficacy (mean 4.6 on 5 point likert scale) among the care team. We also asked participants to rate the educational session utility.

Table 2.

Survey summary statistics for mean 5 point likert responses (5 is most positive) and knowledge expressed as percentages.

	Pre	Post
CT Risk Opinion	4.3	4.8
Knowledge	20%	67%
Willingness to change	-	5
Readiness to change	-	4.6
Educational Session Utility	-	4.9

Semi-structured Interviews

The major points of focus during the interviews with the administrative director, medical director, chief CT technologist, and staff CT technologist were effectiveness of the current content included in the educational session and toolkit as well as identification of issues pertinent to optimizing CT doses that were not addressed by the content of the educational session and toolkit.

Educational Session

When asked about the educational session content, one interviewee highlighted the importance of having facility-specific data in the session, “We all know that dose reduction is important, and in a small hospital that may be where it ends.” Another interviewee suggested that the applicability of the educational session content to the CH is dependent upon the commitment from radiologists for dose optimization: “I have to make sure that the radiologists understand where we are going and I have their buy-in.” Finally, the relatively low volume of CT scans in CHs, particularly for pediatric patients, was a point of discussion among all interviewees (i.e. they lack experience). However, one interviewee noted “just because we're small is not an excuse not to implement good dose reduction strategies.”

Toolkit Usefulness and Refinement

Regarding content in the toolkit, interviewees generally reported that it is a useful resource. One interviewee stated, “Your ABCs of Child Care in CT book is very well written and put together and I think it's really effective to hand that out [immediately following the educational session] while it's still fresh on people's minds… I looked at it that very day, and I found it helpful.” Another interviewee stated:

On a day where we do not have an in-house radiologist and in a situation where time is of the essence, it is not easy to look up CTs online because there are so many images. It is very helpful to have examples of CT scans of specific body regions and dose parameters as a quick guide for our techs. This gives our [radiologists] the information that they need and our techs will be much more comfortable having something to look at. However, because of the low volume of pediatric patients, interviewees suggested that the toolkit contains some information that might not be used: “There are a lot of protocols in there that we will never do.” Another issue that emerged is alignment of protocols between the toolkit and common practice in the facility, for example, weight-based protocols versus age-based protocols.

Dose Optimization Issues Not Covered in the Educational Session or Toolkit

One topic not covered in the educational session or toolkit that we inquired about is discussion of the risks and benefits of CT scans with patients before ordering the study or before performing the study. One interviewee reported, “We try to, yes, but my opinion is most people come in the hospital and that's what they want.” Another interviewee made a similar point about patient's reaction to discussing CT risks and benefits:

We do talk to the patients …and let them know what we are going to be doing, there is radiation involved. …They don't care. They just don't…. We have them come in and say, ‘Well, I need a CT.’ What makes you think you need a CT? You just had one last week.

Another topic involved communication between radiology staff and staff in the Emergency Department:

It's very important and we do that… If they come into the ED, we talk to the ED, and [we say] ‘Is this what you need? What are you looking for? Are you sure you want to order it? Maybe you could do a plain x-ray instead of a CT depending on what you're looking at.’ … As a prime example we had a 12-year-old 2 days ago who came in and they were thinking it was appendicitis. We were hoping to do the ultrasound on him instead. Well, that was not helpful so we ended up having to do the CT. Well, the ED had ordered it with and without [contrast agent]. And it's like, ‘Whoa, no let's not do that. How about we just do it with and that way we can see everything you need to see? There is really no reason to do 2 CT scans on this 12-year-old. … They were good with that, could see he had appendicitis and off to surgery he went. So we are really trying to work with them and they are trying to be very conscientious too and sometimes it's just, ‘Well, this is how we always order it and we thought this is what we needed.’

Finally, interviewees discussed the importance of equipment maintenance and updates:

[T]he key thing that we've got to do … we've got to keep us with current equipment with current updates because again we have a [CT manufacturer] piece of equipment. The major manufacturers are taking this seriously. What we've got to do is make sure that just because we're small it doesn't mean we're not worthy of the upgrades to make this more user friendly and more manageable. I know Joint Commission is probably going to say more on this topic.

Discussion

The Real World

“In an imperfect world, perfection is not instantly available. Railroad safety, for instance, cannot be secured by mechanical devices alone. It is primarily a resultant of care and discipline.”

— Ivy Lee, Human Nature and Railroads, 1915.

Technological advances have played a major role in dose reduction efforts in large academic medical centers [31]. Although technology may be necessary for improvement in practice, it may not be sufficient to change practice, since behavior must be modified across multiple sites of care and professional groups [32, 33]. Therefore, changing practice patterns is complex and such advances are not easily generalizable to the unique context of the CH. In addition to older CT equipment and scanning software, reasons for wide variation in radiation doses from pediatric CT in the CH setting include: 1) complex and counter-intuitive nature of dose-reduction techniques on scanners, 2) lack of education regarding pediatric protocols, 4) absence of specialty trained pediatric CT technologists, 3) lack of a dedicated medical physicist for protocol oversight and quality management, 4) relatively small number of pediatric patients when compared to pediatric-focused facilities, and 5) the possibility that CH radiologists who do not routinely interpret children's CT scans may desire images acquired at higher doses in order to increase diagnostic confidence and to offset the increased noise that generally accompanies lower dose scans [25-26, 34].

To the pressing problem of dose variation in pediatric CT, even large academic medical centers are not immune. In a recent study from the Mallinckrodt Institute of Radiology, St. Louis Children's Hospital, Washington University School of Medicine, Greenwood et al. report substantial variation in radiation exposure per CT examination, even when repeat studies were performed on the same patient [18]. In this project, investigation into such process variation revealed that multiple age and condition-specific pediatric head CT protocols were available on their institution's two scanners, and that there was more variability on weekends and evenings because less experienced technologists were on duty during these times [18]. Interestingly, in our own study, a similar issue regarding off-hours staffing and variability in CT technologists' level of expertise in protocoling patients on the CH's single CT scanner was articulated by the CH imaging director. Dose optimization requires continuous monitoring and periodic feedback to frontline teams.

Our Own Experience

“Experience is that marvelous thing that enables you to recognize a mistake when you make it again.”

— Franklin P. Jones

The impetus for this research was threefold, namely: 1) the lengthy trial-and-error, ongoing experience in a large academic medical center for dose reduction in pediatric CT, 2) the recognition that small hospitals in our own community may struggle with high radiation doses in children's CT, and 3) the belief that we can help them to reduce these doses. Our rationale was that changing practice patterns related to pediatric CT dose reduction is complex, and requires an evidence-based, pragmatic approach to dissemination and implementation [32, 35]. Here we report a pilot study aimed at laying the groundwork for pediatric CT dose reduction in CHs through an academic medical center and CH learning partnership.

Through interactive educational presentations and surveys at a CH, our pilot study revealed an increase in knowledge and stronger opinions about dose optimization in pediatric CT. Additionally, CH stakeholders showed a willingness and readiness to reduce pediatric CT doses. Since our project did not proselytize a target dose for CHs, but instead aimed for dose reduction that was tailored for a CH setting, our goals were uniformly agreeable to key informants. Participants at the CH were interested in monitoring doses and ongoing support and consultation from a larger academic institution. Community hospital leadership sensitive to the importance of CT management and a quality improvement team played a critical role in supporting such an initiative. A focus of future work should be to determine how to refine the toolkit so that users do not perceive it to be too lengthy or inclusive of information that is not pertinent to their setting.

Conclusion

The importance of preserving the human scale in pediatric CT scans cannot be overstated. The tailoring of CT techniques to the clinical questions, small body size and small body parts of children is essential to the practice of dose optimization [36]. Our study revealed that the most challenging aspects for CHs to overcome are opportunities for continuing technologist and radiologist educational workshops, and the limited number of pediatric patients. Although in the CH setting the number of pediatric patients is small, it is an important small number from which to learn and improve.

Take – Home Points.

• Academic medical centers can partner with community hospitals who are eager for help with quality improvement in pediatric CT radiation dose reduction.

• In partnering with the community hospital, we found that they had size specific radiation doses 2 to 3 times higher than those at our center.

• Through interactive educational presentations, an electronic toolkit booklet, and surveys at a community hospital, our pilot study revealed an increase in knowledge and stronger opinions about dose optimization in pediatric CT.

• Results suggested that participants perceived both change commitment (mean 5 on 5 point likert scale) and change efficacy (mean 4.6 on 5 point likert scale) among the care team.

• The relatively small numbers of pediatric patients evaluated in community hospitals provides a unique opportunity for learning and improvement.

The purpose of this study was to develop a partnership between a large academic medical center and a community hospital to develop a toolkit for implementing CT dose reduction, which eventually could be disseminated to community hospitals statewide. Our aims were to 1) assess the acceptability of an interactive educational program and electronic toolkit booklet, 2) conduct a limited test of the efficacy of the toolkit in promoting knowledge and readiness to change, and 3) assess the acceptability and practicality of a collaborative approach to implementing dose-reduction protocols in community hopitals.