Stratied Assessment and Warning Regimen for Comprehensive Risk Management and Improved Iodinated Contrast Media Safety

Objectives: To determine the relationship between stratied assessment and warning (SAW) regimen for risk factors to iodinated contrast media (ICM) and adverse drug reactions (ADR) occurrence. Methods: This retrospective study included patients who underwent enhanced CT examinations in our hospital between January 2014 and March 2016 (conventional assessment group) and between April 2017 and December 2019 (SAW group). Initial risk assessment was performed by clinicians when ordering examinations in both periods. The risk re-assessment was performed by simple question-and-answer onsite just before examination in the conventional assessment period, according to the risk factors that deserve special attention in ICM guidelines. In the SAW period, the re-assessment process prior to examination was moved forward as early as possible post-appointment, and a whole-process comprehensive management integrating risk identication, stratication, early warning and prevention was performed according different risk levels. Chi-square test was performed for comparisons of rates. Results: A total of 120822 cases in the conventional assessment group and 150343 cases in the SAW group were enrolled. The total ADR incidence in the SAW period (414/150343, 0.275%) was lower than that in the conventional assessment period (506/120822, 0.419%, P<0.001), in which the proportion of patients who developed ADR decreased by about 34%. It mainly presented as decrease in mild/moderate reactions (P<0.001) and increase in the proportion of mild ADR patients (P=0.001) in the SAW period. Subgroup analysis showed lower mild/moderate ADR incidence in patients with different risk levels and with different ICM injection parameters (P<0.05) following SAW regimen. Conclusions: SAW regimen was associated with lower mild/moderate ADR incidence and alleviated ADR severity in at-risk patients, which held potential for improved risk management and ICM safety.


Introduction
Iodinated contrast media (ICM) assisted enhanced CT examinations are widely and daily conducted for diagnostic procedures in imaging departments worldwide. Although patients can bene t from their use, ICM possess inherent risk to cause adverse reactions (ADR), with an incidence ranging from 0.2-3.13% [1][2][3][4][5][6][7][8][9]. The clinical manifestations vary from mild unspeci c physiological disturbances to rare severe and potentially life-threatening events [1,10]. The likelihood of ICM-ADR is related to the complicated and combined effects of miscellaneous risk factors, and the occurrence of events follows sporadic and unpredictable pattern. In clinical practice in China, initial risk assessment is usually performed by clinicians at the time of ordering examinations. Due to the limitations of expertise eld and knowledge background, clinicians have limited awareness of ICM risk factors, which may lead to inaccurate assessment results. The risk re-assessment needs to be performed by radiology nurses in the form of simple question-and-answer (yes or no) onsite just before examination, according to the risk factors that deserve special attention outlined in international ICM usage guidelines [11,12]. Inquiring about some complex risk factors may be not detailed or in-depth enough attributed to a tight schedule onsite, making the accurate recognition and strati ed management of risk factors di cult. The lack of objectively and quantitatively standardized assessment programs have led to inconsistent assessment results among different medical staff and medical institutions. Some patients with advanced age and poor selfknowledge ability are unable to communicate accurately and completely describe their medical information. The inaccurate estimation of risk factors may instigate subsequent excessive/insu cient prevention measures, which inevitably led to a series of clinical issues [13]. If high-risk circumstances were recognized when enquiry, no adequate and appropriate preparation could be implemented due to time constraints [14,15]. In this scenario, such a patient population had to be rescheduled for elective examinations, giving rise to delayed radiology diagnosis time and a waste of medical personnel resources [16][17][18]. These issues have drawn increasing attention of the radiology department.
Current ICM guidelines proposed several common types of ICM-ADR risk factors that deserved special attention and highlighted the importance of risk assessment [2,19]. Regrettably, there are no standardized procedures and e cient strategies for strati ed assessment and management of ICM risk factors. Despite that several studies reported the assessment and management of patients at risk [1,18,[20][21][22], their results differed and have not been generally accepted. It is urgently demanded to establish a scienti c, practicable and widely applicable management process for ICM risk factors, which is of clinical importance to reduce the frequency and/or severity of ICM-ADR, and to ensure maximum safety for the patients. From April 2017, our hospital undertook a standardized quality improvement project that involved a strati ed assessment and warning (SAW) regimen for ICM risk factors, which focused on issues concerning ICM-ADR that integrated risk identi cation, strati cation, early warning and prevention, and aimed to improve the ICM safety. With the radiology nurses as the main body, through structured cooperation of the radiological team as well as their collaboration with clinical departments, strati ed assessment and personalized management of risk factors were implemented during the period from patient appointment to ICM injection. This study aimed to determine the relationship between the SAW regimen and acute and delayed ICM-ADR occurrence.

Study Participants
CT scan data was collected and retrospectively analyzed from our hospital, a comprehensive tertiary and grade A medical institution with 2,600 beds that provides all medical and surgical services. This retrospective study was approved by the institutional review board of our hospital. The written informed consent was exempted because this study would not affect the rights of the participants, and all personal data were removed and coded as arbitrary numbers. The research ow chart is shown in Fig. 1. Inclusion criteria: (i) Patients who met the indications and underwent routine CT enhancement from January 2014 to March 2016 (conventional assessment period) and from April 2017 to December 2019 (SAW period) [11,12]. (ii) Patients who had risk factors outlined in international ICM usage guidelines [11,12], but needed CT enhancement for disease diagnosis. (iii) There was no age limit for patients. Exclusion criteria: (i) Patients with incomplete form data lling; (ii) Unconscious patients with unavailable assessment; (iii) Emergency patients with unknown medical history. We allowed a transition period of 12 months (from April 2016 to March 2017) for quality improvement project to fully permeate the examinations, and data from during this period were not analyzed in this study.
Assessment in the conventional assessment group Initial risk assessment was performed by clinicians at the time of ordering examinations. And the risk reassessment was performed on-site just before examination by radiology nurses in the form of simple question-and-answer (yes or no), according to the risk factors that deserve special attention outlined in international ICM guidelines [11,12]. The radiology nurses with over 8 years of work experience asked the patients and their families for the patients' medical history, lled out the conventional assessment form (Table S1, Supplemental Materials), informed the risks for ICM injection, and asked the patients to sign the informed consent form for ICM injection. The patients were closely observed during and after examinations, and abnormal reactions were treated in time and routine hydration were performed. For patients with risk factors [11,12], the radiology nurses reported to the radiologists, and the radiologists should communicate with the clinicians about the individualized risk-bene t ratio of examination, countermand the examination directly, reschedule for elective examinations after clinical treatment if necessary, and consider alternative imaging modalities with comparable diagnostic values.
Strati ed assessment of risk factors in the SAW group Design strati ed assessment form: Continuous quality improvement was carried out on the basis of the conventional assessment. The assessment items for risk factors were optimized and updated, and the degrees of risk factors were strati ed. (i) Determine the risk assessment items: Through previous observation results of hundreds of thousands of CT-enhanced patients [9,20,23], we concluded that patients with asthma, heart disease, hypertension, ICM-ADR history, other allergy history, tumor chemotherapy, and elder patients presented a higher ADR incidence. To ensure the rationality and effectiveness of the screening process, the items in the assessment form were improved by combining our institutional results, international ICM guidelines and the latest literature. (ii) Determine the contents in the assessment form: An expert seminar was held by ve radiologists, ve radiology technicians, and ve radiology nurses. All of them were senior members of the local Contrast Media Safety Committee with deputy senior professional titles and above, and had worked in tertiary hospitals for more than 8 years. The risk factors were discussed item-by-item and were classi ed into different risk levels, including high risk, low risk, and no risk (including unknown risk). Ultimately, 12 screening items were determined in the strati ed assessment form (Table S2, Supplemental Materials).
Move forward the re-assessment process: The re-assessment process prior to examination was moved forward as early as possible post-appointment. The radiology nurses with over 8 years of work experience lled out the strati ed assessment form item-by-item. For patients with unclear information descriptions, their medical histories and related test results were obtained through the medical information system. The assessment of diseases like thyroid diseases, hypertension, heart insu ciency, renal insu ciency, and asthma were based on their nal clinical diagnosis.
ADR record cards and risk warning signboards: (i) ADR record card: The card recorded the date and time for previous ADR, culprit ICM brands, injection dosages and speeds, clinical symptoms, severities, treatment and outcomes. The card could accurately re ect the events that occurred, and was held by the patients themselves and brought to any other hospital for check-up before examination. (ii) Risk warning label: The risk degrees were divided into high-risk, low-risk and no-risk, and the high-risk and low-risk was colored in red and yellow respectively. For patients with multiple risk factors, the most serious risk factor was adopted as the basis for determining the risk degree. The patients' risk levels were carefully determined by one doctor-in-charge and one nurse-in-charge both with over 8 years of work experience. If disagreements occurred, they discussed with clinicians to determine the nal assessment results.

Determine intervention strategies in the SAW group
According to different risk strati cation, corresponding comprehensive intervention was implemented (Table S3,  Different risk signs were labeled on the checklists to remind technicians of formulating personalized examination programs, including adopting lower injection dosages and rates of ICM, switching to another class of ICM for patients with ICM-ADR histories, performing dynamic electrocardiograph monitoring during the examination, closely observing the condition changes, asking whether the patients had abnormal reactions. High-risk patients were followed up after examination by dedicated staff. If any abnormality occurred, the emergency department or residents were informed to immediately start the rescue emergency plan, and follow up was well-performed.

Data documentation and quality control
Detail comparisons of conventional assessment and SAW regimen are shown in Table 1. All patients who underwent enhanced CT examination routinely lled out the conventional assessment form or the updated strati ed assessment form for risk factors ( and some symptoms were equivocal and hard to be differentiated, these two subtypes of ADR were collectively de ned as adverse events. Thyroid toxicity and post-contrast acute kidney injury were not follow-up observed. To ensure the accuracy and consistency of the assessment results, the radiologists, technicians, and nurses received formal trainings in advance, such as thematic lectures, on-site demonstrations, and situational presentations. The training contents include ICM-related risk factors and intervention methods, assessment form usages, clinical information acquirements, informed consents, and how to implement emergency treatments. The evaluation was conducted by dedicated staff, and the data was collected with uni ed standards and checked daily, and the missing items were completed in time. After data registration, two medical staff with over 8 years of work experience checked the original data blind-toblind to ensure the data accuracy and completeness.

Statistical analysis
All variables were descriptively analyzed. Continuous variables were described in terms of mean values and standard deviation. The counting data was presented in terms of frequencies and percentages (%). Chi-square test was performed for rates comparison on SPSS 22.0 (IBM, Chicago, USA), and P < 0.05 was considered statistically signi cant. The rate differences and 95% con dence interval (CI) were estimated using VassarStats website http://vassarstats.net/index.html.

Results
A total of 273437 cases underwent enhanced CT examinations between January 2014 and December 2019, in which "case" was equal to the number of ICM administration. 2272 participants were excluded (Fig. 1), including 722 patients with incomplete form data lling, 643 unconscious patients with unavailable assessment, and 907 emergency patients with unknown medical history. After excluding participants with incomplete follow-up, the complete case analysis consisted of data from 120822 eligible cases in the conventional assessment group (0-104 years old, 58 ± 15 years, 66573 men [55%]) and 150343 eligible cases (0-102 years old, 58 ± 14 years, 83937 men [56%]) in the SAW group (Fig. 1).
Demographic and baseline characteristics of study participants are summarized in Table 2. Note: More than one examination regions might be involved in one enhanced CT examination.
Comparison of the incidence and proportions of ADR with different severities are shown in Fig. 2. The total ADR incidence in the SAW group (414 of 150343 examinations, 0.275%) was lower than that in the conventional assessment group (506 of 120822 examinations, 0.419%, P < 0.001), and the proportion of patients who developed ADR decreased by about 34%. No statistical difference was found in the severe ADR incidence (P = 0.33). The ADR incidence in patients with risk factors is shown in Table 3. Whether patients with no risk or at risk, high-risk or low-risk, single-risk-factor or multi-risk-factor, the SAW group exhibited a lower ADR incidence than that in the conventional assessment group (P < 0.05). The rate difference in high-risk patients (0.515%, 95% CI: 0.262%, 0.768%) was greater than that in low-risk patients (0.124%, 95% CI: 0.005%, 0.244%), and the proportion of ADR in high-risk patients was reduced by about 58%. Comparison of the ADR incidence in patients with different single risk factors are shown in Fig. 3. The ADR incidence in patients with ICM-ADR history, heart disease, hypertension, and advanced age (≥ 70 years) in the SAW group was lower than that in the conventional assessment group respectively (P < 0.05). The proportion of these populations who developed ADR was reduced by about 52%, 60%, 56%, and 41%, respectively. 40 patients who developed ADR in the conventional assessment period and subsequently received ICM again during the SAW period were analyzed (data not shown), and none of them developed recurrent events. The records showed that another non-culprit ICM was used for every patient and no prophylactic medication was performed in these patients. The ADR incidence in patients with different ICM injection parameters are shown in Table 4. Whether isotonic or hypotonic ICM, the SAW group exhibited a lower ADR incidence than that in the conventional assessment group (P < 0.01), which was mainly re ected in mild and moderate ADR (P < 0.05). Whether low or high injection dosages and speeds, the ADR incidence in the SAW group was lower than that in the conventional assessment group (P < 0.05).

Discussion
Current ICM guidelines lack large-scale data base for optimal management of risk factors and effective prevention of ICM-ADR [24,25]. Standardized SAW regimen can standardize the roles and behaviors of medical staff in the radiological team. This may contribute to accurate recognition and strati ed management of risk factors, which is of great signi cance for improving the ICM safety. As a quality improvement project in a large clinical cohort (271165 cases), our study determined the relationship between standardized SAW scheme and ADR occurrence. The total ADR incidence in the SAW group was lower than that in the control group (0.275% vs 0.419%, P < 0.001), with a rate difference of 0.144% (95% CI: 0.10%, 0.19%). The decrease in ADR occurrence mainly presented as a decrease in mild (P < 0.001) and moderate reactions (P < 0.001), and an increase in the proportion of mild ADR patients (P = 0.001).
The relationship between SAW regimen and reduced ADR occurrence Given the critical role of ICM in modern medical imaging and the huge population for enhanced CT examinations worldwide, substantial reduction in the potential risks of ICM administration is of great signi cance to alleviating the social, medical and economic burden [14,15]. The decreased ADR occurrence in the SAW period mainly presented as a higher proportion of mild ADR and a lower proportion of moderate ADR. This is of great importance for alleviating the rescue and disposal loads of medical personnel. Attributed to the extremely low incidence of severe life-threatening anaphylaxis, the Chi-square test result was unreliable due to the very small size of positive events. The relationship between SAW scheme and severe ADR requires larger sample size for veri cation. To investigate relationship between SAW regimen and patient subgroups, patients were divided into no-risk (including unknown-risk) and atrisk based on the assessment results. The greater rate difference of ADR in at-risk patients suggested that the SAW regimen had a more direct and remarkable effect on this population. The effect of SAW scheme on patients with no-risk and at-risk were mainly manifested as decreases in mild and moderate ADR. Further analysis revealed that the ADR incidence in patients with whether high-risk or low-risk, single-riskfactor or multi-risk-factor in the SAW period was lower than the conventional assessment period, respectively. The greater rate difference in high-risk patients suggested that SAW regimen had a more direct and remarkable effect on this population.
The ADR occurrence in patients with various single risk factors was further analyzed. The results showed that patients with ICM-ADR history, heart disease, hypertension, and advanced age (≥ 70 years) in the SAW group had a lower ADR incidence than that in the conventional assessment group respectively (P < 0.05). The proportion of these populations who developed ADR was reduced by about 52%, 60%, 56%, and 41%, respectively. As the possible interference caused by the different basic physical conditions among different patients was eliminated, the aforementioned self-control study results from 40 patients with ICM-ADR history strongly indicated that SAW program was remarkably associated with prevention and control of ADR reoccurrence. While no statistical difference (P > 0.05), the ADR incidence in patients with history of other allergies or cancer was lower. This indicated that some treatment measures (e.g., replacing the culprit ICM, controlling the injection dosages and speeds) might have a certain effect on inhibiting ADR occurrence, which was consistent with previous reports [2,3,19,26,27].
These effect of ICM injection dosages and speeds on ADR occurrence in the conventional assessment group was consistent with reports in the literature [28,29]. Interestingly, there was no statistically signi cant difference in the ADR incidence in patients with high and low injection dosages in the SAW group (P = 0.587), but both were lower than those with low injection dosages in the conventional assessment group. This might suggest that following SAW regimen, the ADR incidence in patients with high dosages could be reduced to a level similar to or below that of patients with low dosages in the conventional assessment group. The injection speed data showed similar results. These results further indicated that the SAW regimen had certain directive signi cance for selecting appropriate ICM parameters. It suggested that if conditions permit in clinical practice, hypotonic ICM could be chosen for replacement for high-risk patients with an isotonic ICM-ADR history, and avoid the injection dose ≥ 100 mL and the injection speed ≥ 5 mL/s whenever possible.
Reasons for lower ADR occurrence following SAW regimen As there were no changes in patient care and quality medical between conventional assessment period and quality improvement period, there was a close relationship between SAW regimen and decreased ADR occurrence. The re-assessment process prior to examination was moved forward as early as possible post-appointment by radiology nurses in the SAW period to guarantee appropriate and adequate preparation of patients prior to examination. Our intervention regimen was not limited to anti-allergy for patients with ICM-ADR histories, it emphasized more on the full cooperation between the radiology department and clinical departments before examination [9,20,23] This study has some limitations. First, this study was not a randomized trial and the confounders were not adjusted when making comparisons, which might exist unrecognized changes and unmeasured differences in patient populations. Prospective multi-center randomized controlled trials will help to further validate the clinical e cacy of SAW regimen. Second, the phenomena that abandoned examinations directly, rescheduled for elective examinations, and selected alternative imaging modalities arising from risk overestimation were signi cantly reduced in the SAW period. However, the exact number of these cases were not documented in details. Furthermore, we focused on the ADR occurrence in patients with different risk levels, but did not further address the relationship between different severities of underlying risk diseases and ADR. Nevertheless, our results showed that following standardized SAW regimen, patients at-risk, especially high-risk patients and the ones with multiple risk factors, had a signi cantly reduced ADR incidence in real-life practice. The cumulative effect of multiple risk factors deserves further veri cation.

Conclusions
In conclusion, the whole-process SAW regimen implemented in a large clinical cohort, built a comprehensive risk management process with high clinical practicality, standardized the roles and behaviors of the radiological team in risk screening, guaranteed the accuracy and reliability of risk assessment, and realized timely identi cation and effective management of risk factors. As a result, SAW regimen was associated with lower mild/moderate ADR incidence and alleviated ADR severity in at-risk patients, which held potential for improved risk management and ICM safety.

List Of Abbreviations
ICM, iodinated contrast media; ADR, adverse drug reactions; SAW, strati ed assessment and warning; CI, con dence interval Declarations Ethics approval and consent to participate: This retrospective study was approved by the institutional review board of our hospital. The written informed consent was exempted because this study would not affect the rights of the participants, and all personal data were removed and coded as arbitrary numbers.

Consent for publication: Not applicable
Availability of data and materials: The datasets during and/or analysed during the current study available from the corresponding author on reasonable request.
Competing interests: The authors declare no con ict of interest.  Study ow diagram.

Figure 2
Comparison of the incidence and proportions of ADR with different severities in the conventional assessment group and SAW group.