The use of low-value imaging: the role of referral practice and access to imaging services in a representative area of Norway

The aim of this study was to examine imaging utilisation in general and the use of CW-examinations in particular, and to investigate the role of referral practice, and access to imaging in terms of travel distance to imaging facilities. The setting for the study was the VVHT catchment area. We found that the overall rate per thousand inhabitants varied from 43 for ultrasound to 269 for CR. CR was the most frequently used modality, which is in line with a previous study using data from 2002 [18]. In our data, CT utilisation constitutes 16% of the total imaging used, similar to national data reporting 15% from 2012 to 2015 [27]. Our study documents a 37% increase in MRI utilisation per 1000 inhabitants (rate from 58 to 92) compared to results from 2002 [18], although our study lacks information about in-patient utilisation. Hence, there is reason to believe that the increase in MRI utilisation from 2002 to 2019 is even higher. However, national data from 2016 states that MRI constitutes 22% of the total out-patient imaging in Norway with a rate of 133 per thousand inhabitants [27], while we found MRI use in 19% and a rate of 92 per thousand inhabitants. This may be due to geographical variations nationally [26].

Knowledge about imaging and referral practice is important in identifying areas to target when trying to reduce low-value imaging. Ringberg and colleagues [7] found an overall high referral rate to secondary care and a striking referral range among Norwegian GPs. Moreover, in a study among 480 radiologists, 76% reported that they had received referrals lacking adequate information during the previous work day, and 63% disagreed with the referrer regarding indication [27]. Hence, referral practice may be an important driver for imaging utilisation in general and for low-value imaging in particular. We found great variations in referral practice, and identified 86 of the 3700 (2.3%) referrers as high-referrers. Interestingly, we found no difference in referral rates between less experienced and experienced referrers.

High-referrers referred patients to approximately 33% of all CW-examinations and 15% of all examinations performed in the catchment area, suggesting that our findings are in line with Hong et al. [8] who found that prior imaging patterns and access to equipment were strong predictors for low-value imaging. They also found that the majority of these referrals came from chiropractors and specialists [8]. Interestingly, none of our high-referrers were chiropractors. Therefore, our study confirms earlier findings indicating that Norwegian chiropractors use imaging more sparsely and adhere to guidelines [30]. However, 20% of high-referrers were specialists (mostly neurologists) who may have high referral rates as an MRI of the head is a standard neurological examination for many indications other than the ones included in CW. It is important to emphasise that we cannot identify the rate of low-value referrals, and it is plausible that many of the specialist referrals were warranted. Nonetheless, variation in referral practice may indicate both over and underuse of imaging.

Furthermore, 0.5% of the referrers were super-referrers who referred patients to 3.6% of the examinations in the catchment area. A total of 59% of the super-referrers were GPs or doctors without specialisation, who have an important role as gatekeepers in the healthcare system. Most super-referrers had more than ten years of experience. Kool and colleagues found that 67% of their responding GPs acknowledged that low-value care practice was regularly provided in general practice [31]. There may be many reasons for high referral rates, such as time pressure and patient-related factors [31, 32], (un)familiarity with guidelines [33], and difficulties in handling professional uncertainties [10]. Several strategies are used to avoid imaging overuse [32]. Interventions targeting referrers, including decision support tools, such as the ESR iGuide, guideline implementation, education, feedback to referrers, using various hand-outs, either alone or in combination, have been tried out. Even though the effect of such examinations varies due to contextual and cultural factors, multi-component interventions that include education seem to be more effective than single- component interventions [34]. Our findings of variation in referral practice may indicate flaws in the GPs' role as gatekeepers. However, some GPs might work part-time at emergency clinics, justifying higher imaging rates. On the other hand, the CW-examinations used to identify super-referrers in this study are not typical trauma examinations, making it difficult to explain the high MRI-rates.

Access to services is a driver for utilisation, where the distance to the provider has proven to play a substantial role in healthcare consumption [17]. Also, geographical variations may mean less equal access to radiological services and poor allocation of health resources [18]. MRI and CT utilisation are reported to be slightly higher in urban than rural municipalities [27]. Interestingly, we found no differences in utilisation in relation to travel distance, but a slightly higher MRI utilisation per inhabitant was observed in the groups with longer travel distances. This may indicate that healthcare services fulfil the goal of providing equal healthcare to all citizens despite varying travel distances. Moreover, our findings demonstrate that people in rural areas are willing to travel for services, which is in line with other studies [17]. Higher examination rates for patients in rural parts of the catchment area may be due to a more elderly population [35], thus needing more health services.

This study has a number of strengths. It is the first study addressing referral practice linked to CW-examinations in a Norwegian context. The data used was derived from both public and private imaging providers, which gives a detailed overview of outpatient radiology utilisation in 2019. Moreover, the identification of high- and super-referrers provides a valuable approach to addressing low-value care utilisation. Information about patient municipalities of residence and their use of radiological services also gives valuable insight into the relationship between healthcare utilisation and accessibility.

One limitation of this study is the lack of data for in-patient examinations and a full dataset on examinations fully paid out of pocket or through health insurance. Furthermore, we lack data on referrers owing their own radiological equipment. For instance, one in five chiropractor clinics possess radiological equipment and one in ten have access to ultrasound [30]. The inclusion of this data would give a more complete picture of the imaging utilisation. As this study is based on registry data, we do not have any information on the clinical question or the percentage of findings with impact on patients’ diagnosis, management, or prognosis. Also, the number of referrals is based on the number of examinations connected to the individual referrer and may not reflect the actual number of referrals as several examination codes may have been registered on one referral. Due to the lack of patient identification, some codes may have been used for the same patient at the same examination session, therefore overestimating the number of referrals. However, the great variation in referral practice suggests that referrers with high referral rates may be a warranted target in reducing low-value imaging. Our findings may not be directly transferable to other settings as the study focussed on one specific area. However, the catchment area covers a substantial proportion (10%) of the Norwegian population and is deemed representative of a Norwegian setting as it consists of both urban and rural settlement areas with a variety of distances to health care services.

In this study, our main analysis focussed on high-volume MRIs as reducing such resource-intensive examinations provides a high potential to free resources for high-value examinations. Further research should include examinations using ionizing radiation due to their implications for radio protection.

留言 (0)

沒有登入
gif