Abstract
Objective
Physicians often use mobile apps for patient care, but few apps are dedicated to pediatric critical care medicine (PCCM). This study developed a mobile app specifically for Pediatric Critical Care Unit (PICU) residents to aid their onboarding process and aimed to assess whether it could enhance their confidence and comfort levels.
Method
From March 2020 to April 2021, 90 residents participated and completed pre- and post-rotation quizzes with 20 questions each. Quiz score differences between the control and app groups were analyzed using t-tests. A survey was also administered at the end of the rotation to compare comfort level and confidence in PCCM knowledge pre- and post-rotation.
Results
Enrollment included 50 residents in the control group and 40 in the app group. The participation rate was 100%, but not all participants completed both quizzes and survey. The app group showed a significantly greater improvement in quiz scores from pre- to post-rotation compared to the control group (increase of 0.23 questions vs 1.67, p = 0.045). However, the two groups had no significant differences in confidence in PCCM knowledge (p = 0.246) or comfort levels (p = 0.776) in the PICU.
Conclusions
This study found no significant difference in confidence levels between the App and control groups at the end of the PICU rotation. However, the App group outperformed the control group in knowledge assessments. Frequent use of the app likely reinforced essential concepts and facilitated adaptation to the PICU service. Overall, the app's positive impact on knowledge and adaptation indicates it is a valuable tool for enhancing medical residents’ educational experiences in busy clinical environments.
Education objectives
To investigate the impact of using a customized mobile app specifically designed for Pediatric Intensive Care Unit (PICU) residents on their confidence and comfort level.
To present the results of the study, including a comparison of quiz scores between the control group and the app group, as well as the differences in comfort levels and confidence in Pediatric Critical Care Medicine (PCCM) knowledge between the two groups
To discuss the findings of the study, including the possible reasons for the greater improvement in quiz scores in the app group and the lack of significant difference in confidence levels and comfort level in the PICU between the groups.
Introduction
In this new era of advanced technology in which smartphones are ubiquitous, increasing numbers of medical providers are utilizing mobile applications (apps) for patient care. Although the number of medical device platforms available for download is expanding rapidly, only a few are devoted to pediatric critical care, as evident from the app stores of major mobile device operating systems. Although there are numerous applications customized for pediatric medicine and adult critical care, the shortage of those specifically designed for pediatric critical care is apparent. In the United States, pediatric and emergency medicine residencies are mandated by the Accreditation Council for Graduate Medical Education (ACGME) to include rotations through the pediatric critical care unit (PICU) as part of their training requirements.1,2 For most of these residents, the PICU rotation presents a significant challenge due to demanding hours and patients with high acuity. The fast-paced environment can be overwhelming, hindering effective information absorption. Additionally, accessing resources and hospital-specific policies can be difficult, further complicating the learning process. In 2017, Wolfe and Unti report that pediatric residents experience more burnout during their PICU rotation, more so compared to other pediatric rotations (65% vs 45%) and screen positive for depression more at the end of the rotation (42% vs 4%). 3
In 2011, a survey conducted by Franko and Tirrell among providers at medical centers recognized by the Accreditation Council for Graduate Medical Education (ACGME) revealed that half of the participants incorporated mobile apps in their clinical practice. The data indicated that medical trainees expressed interest in utilizing apps customized for medical training and promoting self-directed learning. 4 Subsequently, several studies have been published that reinforce the growing interest among medical trainees in leveraging smartphone apps for medical education.5–10 However, few studies exist to quantify the impact of mobile apps on resident education; furthermore, these studies are limited by inadequate or lack of a control group.
Our team created a customized mobile app targeting medical residents rotating through the PICU at Children's Hospital of Michigan (CHM). The app was developed using an online app builder platform (https://thunkable.com/), and includes a range of features such as orientation materials, divisional policies, landmark articles on relevant PICU topics, on-call and lecture schedules, a phone directory with direct dialing links, medication dosing references, pediatric vital signs charts, and medical calculators (Figure 1). The app was exclusively designed for pediatric and emergency residents working in the PICU at CHM located in Detroit, Michigan. We hypothesized that utilizing this innovative app would enhance the medical knowledge, confidence, and comfort level of the residents as a physician in the PICU.
Screenshots of the CHM PICU App: Figure 1A- Table content for the App. Figure 1B- A list of hospital protocols. Figure 1C- A compilation of commonly used medications for pediatric critical care, featuring information on Aminophylline and its dosages for various indications. Figure 1D- A calculation guide for weaning from an intravenous morphine drip or fentanyl drip to oral morphine, scheduled at intervals of every 4 or 6 hours. Figure 1E- Normal vital sign ranges for the pediatric population. Figure 1F- Phone directory for various departments and staff members of the pediatric critical care unit at Children's Hospital of Michigan.
Methods
Study design
From March 2020 to April 2021, we conducted a prospective, observational, survey-based, block-randomized, single-center pre-post exploratory study involving ninety pediatric and emergency medicine residents at the start of their PICU block. Each 4-week block was randomly assigned to either the App or control group using block randomization. This method was crucial in preventing cross-contamination, as allowing some residents within a block access to the mobile app while others did not could compromise the study's integrity. Residents without access might borrow the app or rely on those with access, potentially confounding the results. An online randomizer website (http://www.randomizer.org/) was used to perform the randomization. The control group comprised trainees who did not have access to the mobile app, while the App group included trainees who had access to the customized mobile app “CHM PICU” developed via Thunkable (https://thunkable.com/). The App group residents were provided with a link to download the app, which they were free to use during their rotation. The study cohort consisted of second- and third-year residents specializing in pediatric or emergency medicine. Emergency medicine residents undergo one PICU rotation during their residency, while pediatric residents at Children's Hospital of Michigan complete one PICU rotation in both their second and third years. Trainees who had already participated in the study during their first PICU rotations within the study period were excluded from the analysis. The study excluded medical residents rotating in the PICU as an elective rotation. The Wayne State University Institutional Review Board granted approval for the study via an expedited review process, and all methods adhered to the relevant guidelines and regulations. Informed consent was obtained from all participants.
Knowledge assessments
Upon enrolling in the study, the residents underwent a pre-rotation quiz to assess their knowledge of critical concepts in pediatric critical care. Participation in the study was entirely voluntary, and residents were free to withdraw from the study at any time. At the end of their 4-week PICU rotation, the participants completed a post-rotation quiz. Both quizzes contained 20 questions and had similar content. The range of questions aligned with the content outline for the emergency and critical care domains of the Pediatrics board certifying exam developed by the American Board of Pediatrics. 11 Both the pre- and post-rotation quizzes contained an equal number of questions corresponding to the categories selected from the content outline. All questions were presented in a multiple-choice format. Prior to subject enrollment, we conducted focus groups with PICU residents to ensure that the knowledge assessment quizzes had face validity. Each quiz was labeled for each participant, enabling us to calculate the change in their quiz scores from baseline.
Surveys
We ensured the face validity of our two surveys—one for each study group—by conducting focus groups with PICU residents, fellows, and attendings prior to subject enrollment. The anonymous surveys were completed by participants at the end of their PICU rotation and included demographic questions on age, gender, residency program discipline (pediatrics or emergency medicine), training level (second or third postgraduate year (PGY)), and prior PICU experience. Both surveys also assessed residents’ comfort level and confidence in pediatric critical care medicine (PCCM) knowledge before and after the rotation, with questions presented in a 5-point Likert response format indicating the level of agreement with each statement. 12 The App group survey included 14 additional questions about the CHM PICU app, such as frequency of use, helpfulness in completing tasks, ease of transition into the rotation, and improvement of communication among team members. Survey responses were compared between the two study groups to determine the app's impact on resident comfort level and confidence in PCCM knowledge (Appendices B & C).
Statistical analyses
To assess the primary objective of determining improvement in residents’ PCCM knowledge, we calculated the difference between each resident's paired pre- and post-rotation quiz scores and compared the results between the control and App groups. Secondary objectives included testing the effectiveness of the app in improving residents’ self-efficacy and comfort levels in the PICU.
Demographic variables between the two study groups were compared using non-parametric Fisher's Exact test, while independent sample t-tests were used to compare continuous variables (e.g., assessment knowledge scores) between the groups. Pearson's chi-square test was used to compare categorical and ordinal data. We used adjusted multiple linear and ordinal regression models for secondary outcomes and sensitivity analyses to account for potential confounding factors. All reported p-values were two-tailed, with statistical significance at p-values ≤ 0.05.
Results
Demographics
In the study, a total of 90 residents were enrolled, with 50 assigned to the control group and 40 to the app group. The participation rate was 100%. However, completion rates for both quizzes were lower in the control group, with only 68% (n = 34) of residents completing both quizzes, compared to 92.5% (n = 37) in the app group (p = 0.005). It was challenging to have all residents complete both quizzes because some residents were post-call or actively engaged in clinical duties at the time of the quizzes. Demographic variables, including gender, residency program, level of training, and prior PICU experience, were reported using ratios and percentages (Table 1). No significant differences were found between the two groups on these variables, as determined by Fisher's Exact Test.
Demographic variables and easy to adapt variable for the two study groups.
SA/A: Strongly agree or agree SD/D: Strongly disagree or disagree.
The demographic variables encompassed gender, residency program discipline (Pediatric or Emergency Medicine), training level (second- or third-postgraduate year (PGY) resident), and whether the resident was experiencing their first Pediatric Critical Care Unit (PICU) rotation. The “Easy to Adapt” variable was determined based on residents’ responses to a survey question regarding the ease of adapting to the PICU rotation, rated on a 5-point Likert scale ranging from Strongly Agree to Strongly Disagree. Both the demographic variables and the “Easy to Adapt” variable were obtained from survey data. Statistical analysis was conducted using Fisher's Exact test for 2 × 2 tables and Pearson's Chi-square test for tables extending beyond 2 × 2 table. The results revealed no statistically significant differences between the study groups regarding demographic variables.
Knowledge assessment
For each participant, we calculated the change from the initial quiz score (score range, 0 to 20 [lowest to highest possible score]). Pre- and post-quiz scores and the difference in individual's pre-post test scores were compared at each time point between groups using parametric independent samples t-tests (Table 2). The homogeneity of variance was checked and verified using the Levene test. 13 The baseline knowledge level of the control and App groups was not significantly different (quiz scores of 11.24 vs 11.14; p = 0.87), but the post-test scores between the two groups were statistically different (quiz scores of 11.47 vs. 12.81; p = 0.05), as was the mean pre-post score difference (p = 0.045). The app group's average improvement was 1.67 questions (95% CI [0.91, 2.76]), while the control group's improvement was not statistically significant (95% CI [-0.66, 1.45]).
Mean quiz score between study groups.
A parametric independent samples t-test was utilized to compare quiz scores between the two study groups. The baseline knowledge between the two groups was not significantly different. However, a statistical difference in post-test scores between the two study groups was observed, with a p-value of 0.05.
Similar findings were obtained when comparing the mean pre-post score difference in the control and App groups. The analysis revealed that, on average, the App group's improvement in the number of questions answered correctly was statistically significant, increasing by 1.67 questions from baseline. In contrast, the control group showed an average improvement of only 0.23 questions, which was not statistically significant at a 95% confidence interval.
We conducted sub-group analyses on the mean difference in quiz scores of the App group and found that participants who had previous training in the PICU had a significantly higher improvement in scores (3.83 ± 2.48 questions) compared to those who were training in the PICU for the first time (1.25 ± 2.78 questions) (Appendix A). The mean difference of 2.58 ± 1.22 questions was statistically significant (p = 0.04; 95% CI [0.09, 5.06]). We used multiple linear regression to explore the potential predictors of the mean difference in quiz scores. Gender, residency, level of training, and first-time PICU training were included as potential predictor variables. The best predictor of the mean difference in score was whether the current rotation was the first in the PICU. Specifically, participants with prior PICU training had a mean increase in quiz scores of 2.41 questions compared to those with no prior PICU experience (p = 0.006). The second-best predictor was whether the subject was in the App group, with a mean increase in quiz scores of 1.53 questions (p = 0.024).
Survey data
To examine proportional differences in individual responses between the two groups regarding confidence, comfort, and ease of adapting to the PICU as a resident, we conducted a Pearson's chi-square test. Changes in confidence and comfort levels on a 5-point Likert scale were reported descriptively and graphed in percentages at each point for the study groups (Figure 2). At the beginning and end of the rotation, participants were asked to rate their confidence in pediatric critical care knowledge. In the control group, three subjects (6%) reported a decrease in confidence level by the end of the rotation, whereas none of the App group participants experienced a negative change in confidence. Although there was a trend of increased confidence in pediatric critical care knowledge for the App group, the difference did not reach statistical significance according to the Pearson's chi-square test (p = 0.246). Similarly, there was no statistical difference between the two groups for changes in comfort level in the PICU from the baseline level (p = 0.776).
The study assessed participants’ confidence and comfort levels at the beginning and end of the PICU rotation using a 5-point Likert scale (from 1 being “very confident/comfortable” to 5 being “not confident at all” or “very uncomfortable”). The difference between the baseline and final confidence and comfort levels was plotted in separate frequency bar graphs (Figures 2a and 2b). The results showed a trend toward increased confidence in pediatric critical care medicine knowledge for the App group compared to the control group, with no decrease in confidence reported by any App group participant. Additionally, a higher percentage of the App group experienced a 2- to 3-unit increase in confidence compared to the control group. However, statistical analysis did not reveal a significant difference in confidence levels between the two groups. There was no significant difference in comfort levels between the study groups in completing everyday tasks and providing patient care in the PICU.
When asked about their adaptation to the PICU service as incoming residents on a 5-point Likert scale, the App group reported finding it much easier to adapt than the control group (p = 0.003). Specifically, only 1 out of 40 (2.5%) residents in the App group found it challenging to adapt, compared to 15 out of 50 (30%) residents in the control group.
Ordinal regression models were adjusted for changes in confidence in PCCM knowledge and comfort level after the rotation, using a 5-point Likert scale. The level of training was found to be highly significant for changes in comfort level after the rotation. Compared to PGY3 residents, the log odds for change in comfort scores increased by 1.93 for the PGY2 group. However, no significant difference was observed in the change in comfort score pre- and post-rotation by study groups.
According to the survey responses based on a 5-point Likert scale multiple-choice answer, the App group's survey indicated that the CHM PICU app was frequently used and well-liked, with an 82.5% approval rate. Among App group residents, 90% found the app easy to navigate, 87% believed it improved the onboarding process, 85.5% found it helpful, and 63% reported that it improved communication within the department. Additionally, 70% of residents also learned PCCM topics from the app When asked about the frequency of accessing the app during the rotation, only one resident of the App group fell into the lowest two response categories (“never” and “a couple of times per month”), while the remaining 39 reported using the app from “multiple times per day” to “a couple of times per week.” In response to open-ended questions regarding time saved by using the app, 51.6% (16 residents) reported saving approximately 30 minutes per day, while 16.1% (5 residents) reported saving about an hour per day. The protocol/guideline section was deemed the most useful feature (25, 62.5%), followed by the phone directory (11, 27.5%) (Appendix C).
Discussion
The use of mobile medical apps on smartphones is becoming increasingly popular among physicians for medical education and professional development, but filtering accurate and up-to-date information can be challenging.14–20 Despite tremendous resources hospitals use to create protocols and guidelines, they're inconsistently followed due to lack of awareness or accessibility.21,22 Implementing a smartphone app for a specific clinical rotation can allow residents to access medical information at critical times.
An observational pre-post study found that utilizing a smartphone application with the hospital's antibiogram and treatment guidelines resulted in significant improvement in medical trainees’ knowledge of prescribing antibiotics. 8 However, the study was limited by the absence of randomization in the allocation of participants. The control group was enrolled earlier than the App group, which could have confounded the knowledge assessment scores due to medical trainees’ accumulating experience and knowledge throughout the academic year.
Chreiman et al created a mobile platform to assist medical trainees starting trauma rotation and found that the app helped complete everyday clinical tasks and made the transition to the trauma service easier. 23 The study also included usage frequency and found that the app was accessed an average of seven times per day. However, the study lacked a control group and is limited by small sample sizes. Similarly, many mobile app studies produced from different subspecialties studied the usage pattern but not the app's effectiveness on medical education. For studies that aim to examine the impact of the medical app on the learners, most lack a control group and are limited by small sample sizes.7,24,25 Additionally, no similar study existed for pediatric critical care.
One of our initial concerns was that trainees may over-rely on the smartphone app and unable to recall important medical concepts during knowledge assessment tests. However, our study revealed the opposite result: the app group exhibited improved performance on the post-rotation quiz compared to the control group. This improvement is likely attributed to the app's frequent exposure to pertinent material, which reinforces essential concepts. In contrast, traditional paper and digital manuals are less accessible during emergencies and require more time to reference compared to accessing information on a mobile app. Our data revealed a statistically significant absolute increase of 1.67 out of 20 questions in the App group, indicating an 8.35% improvement in test scores in a dataset with an interquartile range of approximately 13%. This improvement in knowledge could prove invaluable for residents, particularly given their crucial role as primary physician responders in PICU emergencies, where timely and informed decisions are paramount. Furthermore, the discrepancy in quiz score enhancement between the two study groups may be underestimated due to the restricted number and content of the quiz questions.
The statistically significant mean difference of 2.58 ± 1.22 questions in quiz score improvements between residents with prior PICU rotations and those experiencing the PICU for the first time highlights the substantial influence of prior PICU exposure on quiz performance enhancement. The absence of prior exposure to PICU environments may be stressful and overwhelming, hindering the residents’ ability to absorb information effectively. To mitigate these challenges, residents may benefit from attending pediatric critical care orientations or courses prior to commencing their PICU rotation. Despite a higher percentage of subjects with prior PICU experience in the Treatment group, there were no significant demographic differences between the study groups among all test subjects as per Fisher's Exact test and Pearson's chi-square test. This finding is further supported by the adjusted multiple linear regression model. The analysis suggests that the best predictor of the mean difference in scores was whether the current rotation was the first rotation in the PICU. However, it is noteworthy that the second-best predictor was whether the subject belonged to the App group, with a mean increase in quiz scores of 1.53 questions (p = 0.024). This suggests that the use of the mobile app had a statistically significant impact on quiz performance.
While there was no significant difference in self-perceived knowledge of PCCM between the two study groups, residents in the App group demonstrated a noticeable trend of improvement in quiz scores. Overall, the findings suggest that this app effectively enhanced the PCCM knowledge of residents, although the full extent of its impact may not have been fully realized.
We hope to improve the PICU experience of residents and reduce burnout and depression by using the CHM PICU app. However, no significant difference was found in the increase of Comfort Level in PICU between the two groups, indicating that the app may not be effective in preventing burnout and depression among residents. Nonetheless, the app was well-received by its users. Most participants agreed that the app facilitated their transition to the new service and helped them perform daily clinical duties in the PICU. Compared to the control group, the App group found it much easier to adapt to the PICU service as incoming residents (p = 0.003; Table 1). Approximately half of the app users estimated that 30 minutes were saved daily using the app.
This study's strength lies in its prospective design with block randomization, which prevented residents in the same rotation from sharing the app. Additionally, this study is the first of its kind to investigate the effectiveness of a customized mobile app in improving resident medical knowledge and experience in a pediatric critical care unit.
There are several limitations to consider in this project. First, since the post-block quiz was conducted immediately at the end of the rotation, it is unclear whether the App group's knowledge improvement is maintained long-term. The second limitation is the imbalance in the percentage of residents who completed both quizzes in each study group, raising the possibility of selection bias. We addressed this confounding variable by performing adjusted multiple linear regression modeling and subgroup analyses for test score improvement.
The third limitation of this study is the discrepancy in the number of eligible participants, with 50 in the control group and 40 in the App group. This variation can be attributed to several factors. First, participant availability: some residents who started the rotation did not complete the end-of-block survey due to post-call duties, clinical responsibilities, or personal reasons. Second, assignment variability: despite the implementation of block randomization, the actual number of participants varied among blocks due to the assignment processes utilized by different residency programs. Third, time constraints: following approval from the Wayne State University IRB, recruitment began in March 2020 and was conducted over a 14-month period to allow adequate time for data analysis and presentation to the Scholarly Oversight Committee before the primary investigator's fellowship graduation. However, the study employed appropriate statistical methods, including adjusted multiple linear regression and sensitivity analysis, to address potential confounding variables and differences in group sizes, thereby supporting the validity of the results despite the discrepancies in participant numbers.
Another limitation is the validity testing of the quizzes and questionnaires; both were designed specifically for this study. Before launching the study, only face validation on the quizzes and surveys was performed by conducting informal focus group sessions with our PICU residents, fellows, and attendings. Other forms of validation were not performed. Lastly, the study is restricted by its small sample size, the sample size and power analysis were not performed as the study period was limited to one year. In future studies, we would like to prolong the study duration and conduct a second post-rotation quiz a couple of months after the rotation to evaluate the sustainability of knowledge improvement in the App group over an extended period.
Conclusion
Although there was no statistically significant difference in the improvement in confidence levels between the two study groups at the end of the PICU block, the App group outperformed the control group on the knowledge assessment test. This could be explained by the App group's frequent exposure to relevant material through the app, which reinforced their recall of essential concepts. Moreover, the customized app helped incoming residents adapt to the PICU service more easily. Introducing a service-specific mobile app is an innovative and novel approach to delivering medical education.
Considering residents’ busy schedules and fast-paced work environments, an intuitive platform like this may improve residents’ self-efficacy and experience while on clinical rotations. While the app did not lead to a statistically significant difference in confidence levels, the significant improvement in knowledge and ease of adaptation to the PICU service suggest that the app can be a valuable tool in enhancing the overall educational experience of medical residents
Supplemental Material
sj-docx-1-mde-10.1177_23821205241275357 - Supplemental material for Customized Mobile App for Residents Rotating Through Pediatric Critical Care Unit
Supplemental material, sj-docx-1-mde-10.1177_23821205241275357 for Customized Mobile App for Residents Rotating Through Pediatric Critical Care Unit by Yu Shan Tseng, Ronald Thomas and Ajit Sarnaik in Journal of Medical Education and Curricular Development
Supplemental Material
sj-docx-2-mde-10.1177_23821205241275357 - Supplemental material for Customized Mobile App for Residents Rotating Through Pediatric Critical Care Unit
Supplemental material, sj-docx-2-mde-10.1177_23821205241275357 for Customized Mobile App for Residents Rotating Through Pediatric Critical Care Unit by Yu Shan Tseng, Ronald Thomas and Ajit Sarnaik in Journal of Medical Education and Curricular Development
Supplemental Material
sj-docx-3-mde-10.1177_23821205241275357 - Supplemental material for Customized Mobile App for Residents Rotating Through Pediatric Critical Care Unit
Supplemental material, sj-docx-3-mde-10.1177_23821205241275357 for Customized Mobile App for Residents Rotating Through Pediatric Critical Care Unit by Yu Shan Tseng, Ronald Thomas and Ajit Sarnaik in Journal of Medical Education and Curricular Development
Supplemental Material
sj-docx-4-mde-10.1177_23821205241275357 - Supplemental material for Customized Mobile App for Residents Rotating Through Pediatric Critical Care Unit
Supplemental material, sj-docx-4-mde-10.1177_23821205241275357 for Customized Mobile App for Residents Rotating Through Pediatric Critical Care Unit by Yu Shan Tseng, Ronald Thomas and Ajit Sarnaik in Journal of Medical Education and Curricular Development
Footnotes
Acknowledgements
We used the SQUIRE checklist when writing our report [Ogrinc G, Davies L, Goodman D, Batalden P, Davidoff F, Stevens D. SQUIRE 2.0 (Standards for QUality Improvement Reporting Excellence): revised publication guidelines from a detailed consensus process.]
Authors’ contributions
Yu Shan Tseng, MD, contributed to the conception and design of the study, contributed to the acquisition and interpretation of the data, drafted the manuscript and prepared the figures, and critically revised the manuscript. Ronald Thomas, PhD, contributed to the analysis of the data and critically revised the manuscript. Ajit Sarnaik, MD, contributed to the conception and design of the study and critically revised the manuscript. All authors reviewed the manuscript, gave final approval, and agreed to be accountable for all aspects of work ensuring integrity and accuracy.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethics approval and consent to participate
Subjects from both the control and treatment groups were recruited at the onset of their PICU rotation, provided with an information sheet detailing informed consent. By completing the pre-block quiz attached to the information sheet, subjects indicated their agreement to participate in the study. The information sheet emphasized that participation was entirely voluntary and that subjects retained the option to withdraw from the study at any point.
The Wayne State University Institutional Review Board approved this study under an expedited review process. All methods were carried out in accordance with relevant guidelines and regulations. Inform consent was obtained from all subjects.
Ethics Approval Date: 09/23/2019.
Funding
The authors were awarded funding by the Sarnaik Endowment Grant
Sarnaik Endowment Grant,
Availability of data and materials
Data generated or analyzed during this study are included in this published article and its supplementary information files.
Supplemental material
Supplemental material for this article is available online.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
