Children with isolated unilateral multicystic dysplastic kidney (MCDK) or congenital solitary kidney (CSK) undergo serial renal ultrasonography with variable frequency until they are transitioned to adult care. A growing body of literature suggests the value of frequent ultrasonography in this population is limited, providing no benefit to overall outcomes. Despite emerging evidence, ultrasound remains overused, resulting in avoidable health care expenditures and unnecessary use of resources. With our initiative, we aimed to improve quality of care by reducing avoidable ultrasounds in these children.
This was a single-center, prospective, interrupted time series of children <18 years with ultrasound-confirmed isolated unilateral MCDK or CSK in the outpatient nephrology clinic to evaluate the effect of a decision-making algorithm on the proportion of children receiving an avoidable ultrasound. An algorithm depicting a consensus, evidence-based protocol for managing pediatric MCDK or CSK was refined through content expert feedback and usability testing to standardize frequency of ultrasonography. Ultrasounds were deemed necessary after birth, at 6 months, and at 2, 5, 10, and 15 years. Differences pre- and postintervention were determined by using a U chart and t and F tests for significance.
The algorithm resulted in a 47% reduction (P < .001) in the proportion of avoidable ultrasounds ordered in children with MCDK and CSK. This reduction was sustainable over a 6-month period and would result in at least $46 000 annual savings.
Introduction of a clinical decision-making algorithm was associated with a reduction in avoidable ultrasound testing. Improving adherence across providers may allow for an even more pronounced reduction.
Unilateral multicystic dysplastic kidney (MCDK) is one of the most common congenital anomalies of the kidney and urinary tract, with an incidence of 1:4300 live births.1,2 For children with unilateral MCDK, and a contralateral kidney that is structurally normal with compensatory hypertrophy, the long-term prognosis is largely excellent.3
MCDK is typically diagnosed on prenatal ultrasound4 and confirmed postnatally with ultrasonography.5 Although there is little debate surrounding the need for ultrasound imaging to diagnose MCDK, there are no consensus- and evidence-based guidelines for the long-term follow-up of these children.6 Literature published as early as the 1970s indicated that after confirmation that the MCDK is isolated and demonstrates classic presentation, further intervention is not warranted,7 but the appropriate length of time for surveillance remains controversial.8 Although current evidence supports conservative management with limited ultrasound follow-up,9,10 there is significant practice variation among providers (Table 1), with some recommending as many as 32 ultrasounds before 8 years of age.11
Existing Algorithms for the Management of MCDK or CSK in Children
| Center | Country | Author | Year | Ultrasound Protocol |
|---|---|---|---|---|
| Monash Children’s Hospital | Australia | Chang et al10 | 2018 | At 3 mo (or earlier if antenatal scan was done before 30 wk), 12–24 mo to confirm contralateral kidney growth and involution of MCDK, and postpuberty10 |
| University of Michigan | United States | Eickmeyer et al9 | 2014 | At birth; if complete involution and normal contralateral kidney, discharge from the hospital to primary care provider for follow-up of blood pressure; if persistent MCDK but normal contralateral kidney, discharge from the hospital to primary care provider for follow-up of blood pressure and ultrasound at 1 year of age to confirm stability or involution of MCDK and refer to specialist if ultrasound results are concerning9 |
| Zeynep Kamil Maternity and Children Training and Research Hospital | Turkey | Moralioğlu et al25 | 2014 | Birth, 3 mo, 6 mo, then every 6 months until 24 mo, then annually25 |
| Children’s Hospital of Winnipeg | Canada | Personal communication (A. Dart, MD, MSc; 2019) | 2010 | At 0–1 mo, 3–6 mo, and 2 y; if normal function and compensatory hypertrophy of the contralateral kidney at 2 y, discharge from hospital to primary care for blood pressure and urinalysis; if incomplete compensatory hypertrophy at 2 y, continue ultrasonography every 2–3 y |
| University of California | United States | Cambio et al1 | 2008 | Every 3–6 mo for the first year, every 6 mo for the second year, and yearly thereafter until 4 y of age1 |
| Nottingham University Hospitals and The Ohio State University College of Medicine | United Kingdom and United States | Sukthankar et al,27 Aslam et al,26 and Hains et al28 | 2000, 2006, and 2008 | At birth, 4 weeks, and age 2 y, 5 y, and 10 y,26–28 then discharge from the hospital to primary care provider for blood pressure and urinalysis annually29 |
| Yale University School of Medicine | United States | Weinstein et al30 | 2008 | At 1 y, 2 y, and then every 2–3 y to assess for involution and contralateral growth until contralateral kidney has reached at least the 95th percentile and the affected kidney has begun or completed involution30 |
| Dokkyo University School of Medicine | Japan | Nakai et al31 | 2003 | Every 3–6 mo for the first year of life, then every 6–12 mo until age 5 y, then annually31 |
| University of Bonn | Germany | Rudnik-Schöneborn et al32 | 1998 | At 3 mo after diagnosis and then yearly32 |
| Royal Liverpool Children’s Hospital | United Kingdom | Rickwood et al33 | 1992 | Every 6 mo until 2 y of age, then annually to 5 y of age with discharge from the hospital to primary care for blood pressure annually33 |
| Center | Country | Author | Year | Ultrasound Protocol |
|---|---|---|---|---|
| Monash Children’s Hospital | Australia | Chang et al10 | 2018 | At 3 mo (or earlier if antenatal scan was done before 30 wk), 12–24 mo to confirm contralateral kidney growth and involution of MCDK, and postpuberty10 |
| University of Michigan | United States | Eickmeyer et al9 | 2014 | At birth; if complete involution and normal contralateral kidney, discharge from the hospital to primary care provider for follow-up of blood pressure; if persistent MCDK but normal contralateral kidney, discharge from the hospital to primary care provider for follow-up of blood pressure and ultrasound at 1 year of age to confirm stability or involution of MCDK and refer to specialist if ultrasound results are concerning9 |
| Zeynep Kamil Maternity and Children Training and Research Hospital | Turkey | Moralioğlu et al25 | 2014 | Birth, 3 mo, 6 mo, then every 6 months until 24 mo, then annually25 |
| Children’s Hospital of Winnipeg | Canada | Personal communication (A. Dart, MD, MSc; 2019) | 2010 | At 0–1 mo, 3–6 mo, and 2 y; if normal function and compensatory hypertrophy of the contralateral kidney at 2 y, discharge from hospital to primary care for blood pressure and urinalysis; if incomplete compensatory hypertrophy at 2 y, continue ultrasonography every 2–3 y |
| University of California | United States | Cambio et al1 | 2008 | Every 3–6 mo for the first year, every 6 mo for the second year, and yearly thereafter until 4 y of age1 |
| Nottingham University Hospitals and The Ohio State University College of Medicine | United Kingdom and United States | Sukthankar et al,27 Aslam et al,26 and Hains et al28 | 2000, 2006, and 2008 | At birth, 4 weeks, and age 2 y, 5 y, and 10 y,26–28 then discharge from the hospital to primary care provider for blood pressure and urinalysis annually29 |
| Yale University School of Medicine | United States | Weinstein et al30 | 2008 | At 1 y, 2 y, and then every 2–3 y to assess for involution and contralateral growth until contralateral kidney has reached at least the 95th percentile and the affected kidney has begun or completed involution30 |
| Dokkyo University School of Medicine | Japan | Nakai et al31 | 2003 | Every 3–6 mo for the first year of life, then every 6–12 mo until age 5 y, then annually31 |
| University of Bonn | Germany | Rudnik-Schöneborn et al32 | 1998 | At 3 mo after diagnosis and then yearly32 |
| Royal Liverpool Children’s Hospital | United Kingdom | Rickwood et al33 | 1992 | Every 6 mo until 2 y of age, then annually to 5 y of age with discharge from the hospital to primary care for blood pressure annually33 |
Congenital solitary kidney (CSK), with a prevalence of 1:1300 children, is largely asymptomatic and follows a similar clinical course with regard to kidney function.12
Global estimates of the costs of surveillance ultrasounds for MCDK throughout childhood are $1.2 to $10 million (US dollar [USD] equivalent) per 1000 affected children, with variations based on geographic location and frequency of assessment.9,11,13,14
Locally at The Hospital for Sick Children (hereafter referred to as SickKids), a pediatric quaternary care center in Toronto, Canada, up to $6.7 million (USD equivalent) is spent per 1000 children with MCDK by using existing practices from birth to age 18 years. Expenditures can be substantially reduced by aligning practices with emerging evidence.
Unnecessary imaging may also lead to incidental findings with no clinical relevance, resulting in the potential for further investigative imaging and unnecessary anxiety for the child and their family.
Reducing unnecessary ultrasounds may reduce institutional expenditures, increase availability of resources for other patients, reduce patient and family expenditures, increase societal productivity by allowing patients and families time to attend school or work, and reduce anxiety surrounding avoidable hospital appointments.
Common factors contributing to overuse of tests include (1) patient expectations and avoidance of conversations regarding appropriate tests with patients,15 (2) fear of missing a diagnosis and associated medicolegal concerns,1,16–18 (3) having been trained to survey for trends over time,16 and (4) a lack of standardized clinical practice guidelines.16,19–21
Lack of standardized protocols and consequent reliance on individual experience has contributed to the high variability in practices among providers in ordering unnecessary tests.21
Whereas there have been no publications to date on reducing unnecessary testing for MCDK or CSK, interventions have been reported in other populations. These have been focused on managing patient expectations,15,22 dispelling false perceptions of risks,15,22 changing physician attitudes,20 and standardizing care protocols through decision-making algorithms.16 Because of its reported effectiveness at reducing unnecessary testing in previous literature, our study was focused on the latter.
With this quality improvement (QI) initiative, we aimed to reduce the proportion of avoidable ultrasounds ordered in children with isolated unilateral MCDK or CSK at SickKids by 80% between September 2019 and June 2020.
Methods
Study Design
This was a single-center, prospective interrupted time series study used to assess the effect of a clinical decision-making algorithm on ultrasound ordering for children with isolated unilateral MCDK or CSK. Patients were included if they had a diagnosis of MCDK or CSK confirmed on renal ultrasound and were followed in the pediatric nephrology outpatient clinic. Patients with CSK were included because of their similarity in long-term outcomes and management to MCDK. Patients were excluded if they had a contralateral kidney or urinary tract abnormality, solitary kidney due to previous nephrectomy, or history of recurrent urinary tract infections warranting additional imaging. The study was completed in accordance with the Model for Improvement, a scientific framework for QI initiatives, focusing on small-scale tests of change to drive continuous improvements.
Context and Setting
The SickKids outpatient nephrology clinic sees ∼400 children with MCDK or CSK annually. Patients are typically first seen 0 to 1 month postnatally, at which point an ultrasound is performed to confirm the prenatal diagnosis. Historically, many children are subsequently monitored via ultrasound at 3, 6, and 12 months and annually or biannually thereafter until transitioned to adult primary care by age 18 years (ie, up to a total of 21 ultrasounds during childhood).
This effort is particularly relevant in the context of the overarching aims of Choosing Wisely, an international campaign initiated by the American Board of Internal Medicine (2012) encouraging physicians and patients to reduce health care tests that consume resources without providing value to patients.16
Diagnostic Methods and Results
All of the 376 patients with MCDK or CSK seen in the SickKids nephrology clinic in the year before this project initiation received at least 1 avoidable ultrasound despite compensatory hypertrophy of the contralateral normal-appearing kidney.
A combination of diagnostic tools was used with stakeholders to interrogate the reasons for the high frequency of surveillance ultrasounds ordered in children with MCDK and CSK. Methods included (1) process mapping to understand the steps involved in an ultrasound appointment; (2) cause and effect with the 5 whys analysis (an iterative, interrogative technique) to understand the root causes of the problem; (3) nominal group technique with dot voting to understand the most frequent contributors to the problem, confirmed with a check sheet to outline the most probable causes; and (4) a Pareto chart to determine the critical root causes responsible for the variability in ultrasound ordering practices.
The stakeholder group cited the following critical root causes: (1) physicians being trained about the importance of tracking disease longitudinally, (2) lack of a local protocol for the management of children with MCDK or CSK, and (3) forgetting to adhere to updated follow-up guidelines.
Intervention
Our intervention was chosen to target the high and variable frequency of avoidable ultrasounds ordered in children with MCDK and CSK and involved developing and implementing a local algorithm to aid in clinician decision-making around follow-up management of this population. Standardization was intended to reduce variability in the frequency of ultrasound ordering and increase adherence to published guidelines for imaging follow-up.
Development and Implementation of the Algorithm
A visual algorithm outlining the protocol for managing children with MCDK or CSK (Fig 1) was drafted by using information from published pediatric guidelines (Table 1) and stakeholder experience. Information on follow-up management, including ultrasonography and annual blood pressure, urinalysis, and creatinine measurements, was included to ensure that patients would be adequately monitored for (1) interval growth and compensatory hypertrophy of the contralateral kidney, (2) anticipated involution of the MCDK, and (3) hypertension, proteinuria, and reduced kidney function. More frequent ultrasounds were included up to 2 years of age because kidney growth parallels overall growth in children. Although not commonly cited in previous literature, a late time point at 15 years was also included to ensure that patients would be assessed after puberty to provide anticipatory guidance before transitioning to adult care with regard to healthy lifestyle and the need for continued monitoring (urinalysis, blood pressure, and kidney function) in adulthood.23,24
Clinical decision-making algorithm for the management of children with MCDK and CSK. a Complex features: hydronephrosis or hydroureter, ectopic or pelvic kidney, abnormal echogenicity, cysts, and/or suspected genetic syndrome. eGFR, estimated glomerular filtration rate.
Clinical decision-making algorithm for the management of children with MCDK and CSK. a Complex features: hydronephrosis or hydroureter, ectopic or pelvic kidney, abnormal echogenicity, cysts, and/or suspected genetic syndrome. eGFR, estimated glomerular filtration rate.
The content was initially reviewed by a single pediatric nephrologist and subsequently revised. Subject matter experts (N = 12 pediatric nephrologists) were engaged to provide feedback on the accuracy, completeness, and clarity of the algorithm, resulting in further revisions. An end-user group consisting of pediatric nephrology staff physicians and trainees (N = 7) was then asked to provide feedback on the style and format, and minor adjustments were made. The algorithm was tested with a single end-user staff physician on a small group of patients for usability before being fully implemented across all patients with MCDK and CSK by all treating physicians. Paper copies (Fig 1) were posted in all clinical areas in which ultrasounds are ordered. An electronic PDF copy was posted on the local shared drive, to which all members of the division have access.
Cost Analysis
The local cost of an avoidable ultrasound appointment was determined through activity-based costing methods. Representative members from each role type (eg, nursing, clinic clerk, etc) were asked to describe the activities involved in an ultrasound appointment. Times for each step were prospectively observed through 5 appointments, and mean times for each activity were calculated. The median of the salary range for each role was obtained through human resources and used to determine the cost per hour for each role type. Costs associated with supplies and tests used during these appointments were found through the billing review function in the electronic medical record system, Epic.
Family of Measures
We used a family of measures (outcome, process, and balancing) throughout this project.
Outcome Measure
The outcome measure was the proportion of avoidable ultrasounds ordered in patients with MCDK or CSK. An ultrasound was considered avoidable if it fell outside of the following time periods: postnatal (within 1 month of birth), 6 months, and 2, 5, 10, and 15 years. These periods were determined to be clinically necessary for directing management on the basis of consensus opinion of the subject matter experts (pediatric nephrologists, radiologists, and community pediatricians) (N = 12) consulted for this project, as well as published reports.10
Process Measure
To enhance measurement feasibility and response rate, the self-reported proportion of patients for whom a provider correctly applied the algorithm (ie, ordered the ultrasound at the appropriate time point using the algorithm) was used as the process measure.
Balancing Measures
Two balancing measures were employed: (1) number of re-referrals to the nephrology clinic within 30 days of discharge (to ensure that reducing ultrasounds and thereby discharging patients earlier was not resulting in patients being re-referred for assessment) and (2) the mean number of renal imaging tests excluding ultrasounds ordered (to ensure that by reducing ultrasounds we were not inadvertently causing an increase in other tests).
Evaluation
Outcome data were obtained through manual abstraction from Epic. Six months of historical data before project initiation were collected retrospectively (March–August 2019). Throughout the project (September 2019–June 2020), data were prospectively collected after each clinic. Of those meeting eligibility for inclusion from each clinic, the proportion for whom avoidable ultrasounds were ordered was determined by reviewing each chart to assess whether an ultrasound was ordered outside of the recommended time periods.
The process measure was assessed by using the mean of each provider’s self-reported adherence to the algorithm, and then confirming that the ultrasound was appropriately ordered by manually reviewing each patient’s orders.
Data on the balancing measure of re-referrals within 30 days were determined by checking the referral history for each patient in Epic. The number of nonultrasound imaging tests ordered was assessed via chart review.
Statistical Analysis
This was a time series study. Statistical process control U charts (QI Macros 2019, Microsoft Excel; Microsoft, Redmond, WA) were used to determine the effect of the algorithm on the proportion of patients with avoidable ultrasounds ordered over time, including whether the change over time was significant, due to special versus common cause variation (control limits set at P < .05), and sustainable. Limits were recalculated once a sustained change was observed on the U chart. Student’s t tests and F tests (IBM SPSS Statistics, IBM Corporation, version 21) were used to determine if the difference in the proportion of ultrasounds ordered, as well as the variance pre- and postintervention, was statistically significant, respectively.
Ethical Considerations
This study was reviewed and approved by the SickKids QI projects review process.
Results
Outcome Measure
In total, 400 patients with MCDK or CSK were seen in the nephrology clinic at SickKids between March 2019 and June 2020. Of these, 241 were included in the final analysis (Fig 2). During the baseline period (March to September 2019), the process was stable, with variation attributed only to common causes; 111 children were included, and 70% completed avoidable ultrasounds. During the 9-month intervention period (September 2019 to June 2020), 130 children were included. During this time, implementation of the algorithm resulted in a significant reduction in the percentage of children completing avoidable ultrasounds (37% of children receiving avoidable ultrasounds [Fig 3], resulting in a 47% reduction from the baseline rate; P < .001 [Fig 4]). Variance was unchanged (F [32,22] = 1.69; P = .20).
Proportion of patients with MCDK and CSK with inappropriate ultrasounds (U chart). CL, centerline; COVID-19, coronavirus disease 2019; UCL, upper control limit.
Proportion of patients with MCDK and CSK with inappropriate ultrasounds (U chart). CL, centerline; COVID-19, coronavirus disease 2019; UCL, upper control limit.
Proportion of patients with MCDK or CSK with inappropriate ultrasounds pre- and postimplementation of the algorithm. *** P < .001
Proportion of patients with MCDK or CSK with inappropriate ultrasounds pre- and postimplementation of the algorithm. *** P < .001
Process Measure
The process measure revealed incomplete uptake of the algorithm intervention across all providers (74%). After further analysis, 2 providers were responsible for the majority of cases in which the algorithm was not applied.
Balancing Measure
There were no re-referrals within 30 days for patients discharged from the hospital, and no nonultrasound renal imaging tests were ordered for included patients over the study period.
Cost Analysis
After accounting for staff time and testing costs, each ultrasound appointment was found to cost ∼315 USD.
Discussion
Interpretation of Results
Implementation of a clinical decision-making algorithm led to a significant reduction (47%; P < .001) in the proportion of patients with MCDK and CSK being managed with avoidable surveillance ultrasounds over the study period. Notable improvements, as evidenced by special cause variation, began when study discussions commenced in September 2020, suggesting a Hawthorne effect. However, gains in improvement continued over the remainder of the study period and were sustainable over the long-term (December 2019 to June 2020).
Our goal was to achieve an 80% reduction in avoidable ultrasounds in children with MCDK and CSK, amounting to 13.6% of children with MCDK or CSK receiving avoidable ultrasounds. To date, we have achieved a 47% reduction (37% of children receiving avoidable ultrasounds). Importantly, these improvements occurred even with incomplete uptake of the intervention. This was likely because of the fact that providers were not explicitly asked to exclusively follow the algorithm for all patients; some providers reported using it selectively. By improving uptake across the 2 providers who least used the algorithm, there is the achievable potential to reduce avoidable ultrasounds further. Doing so would achieve an 87% reduction with only 9% of children with MCDK or CSK receiving avoidable ultrasounds.
Notably, coronavirus disease 2019 began spreading in Ontario in January 2020 and a pandemic was declared in March 2020. Outpatient nephrology clinic visits were reduced from March 2020, with many visits completed virtually, resulting in fewer patients with MCDK and CSK being seen overall. The use of a U chart for this study allowed us to account for the variability in the volume of patients seen in-person.
Comparison With Literature
Consistent with previous studies, improved communication between providers and patients is critical for reducing unnecessary testing.22 Use of an algorithm may also help to facilitate communication between providers and families around the expected follow-up for these children.
A common issue reported with implementing algorithms is individual physician practice variability and the need for a consensus on which follow-up protocol to adopt. Our intervention balanced evidence from the literature with the comfort level of the physician group based on their collective experience in a stepwise approach to ensure that the intervention was followed and sustainable.
Impact
Appointments for MCDK or CSK at our institution occur up to 21 times from birth to age 18 years using historic practices; 7 of these (47%; or the equivalent of 1 every 2.5 years from birth to 18 years of age) can be eliminated through implementation of the algorithm, amounting to a reduction of 46 200 USD/year across all patients. With minimal up-front development and maintenance costs, the algorithm is expected to generate an overall cost savings of ∼45 800 USD/year. With fewer patients with MCDK and CSK booked for follow-up ultrasound appointments, administrative, nursing, nephrologist, ultrasonographer, and radiologist time allocated to these patients, as well as supplies and tests needed for urinalysis that is always done during these appointments, will be reduced. To properly realize the cost savings, the algorithm would need to be successfully and consistently applied by all providers across all eligible patients. This includes using the algorithm for eligible patients seen outside of nephrology (eg, urology clinic).
Total savings from implementing the algorithm extend beyond the institution. For patients and families, less time and less money are spent on transportation, parking, or food while in the hospital. Children avoid missing school and parents avoid missing work to attend appointments, which garners a substantial societal benefit. The cost savings outlined above therefore underrepresent the total potential savings.
The ease of use, significant initial results despite incomplete uptake, low maintenance costs, and substantial cost savings to the institution are factors that contribute to the long-term sustainability of the algorithm.
Limitations
This was a single-center, nonrandomized, before-and-after study. In this study, we did not employ the use of a control group; thus, alternative explanations for the observed changes in the proportion of patients completing avoidable ultrasounds are possible, and one cannot say definitively that the observed differences were due to implementation of the algorithm. The length of historical data to establish the baseline rate, temporal association of the change in outcome with the implementation of the algorithm, and use of a process measure to track adherence to the intervention lend credibility to the intervention having been responsible for the changes observed in the outcome.
This study employed the use of a self-report process measure, which may be subject to social desirability and recall biases. Moving forward with this intervention will require that the process measure is tracked continuously to ensure consistency with uptake across providers and patients.
A 30-day re-referral period was chosen as a balancing measure to capture patients with immediate concerns; however, concerns resulting in re-referral at any point during childhood will be important to understand the long-term consequences of reducing surveillance ultrasounds.
Long-term kidney function was not measured because serial serum creatinine measurements were not available during the study period. However, creatinine is expected to be normal in young children with normal-appearing solitary kidney that shows interval growth and compensatory hypertrophy, with no proteinuria and normal blood pressure. Monitoring of urinalysis and blood pressure are included at every follow-up visit in our algorithm.
Conclusions
Unilateral MCDK and CSK are common congenital renal anomalies and carry an excellent prognosis for children, provided the contralateral kidney is normal. Lack of standardized guidelines for surveillance ultrasonography have contributed to inconsistent resource use, which may negatively impact patients, families, and health care institutions. With the implementation of a cost-effective, evidence-based, locally approved clinical decision-making algorithm, quality of care for children with MCDK or CSK may be improved by reducing avoidable ultrasounds.
Future goals are to improve adherence by tracking fidelity continuously to understand factors contributing to nonadherence, to implement the algorithm into Epic by using best practice alerts to advise users as to the recommended time points for ultrasound ordering, to track kidney function and re-referrals to nephrology after discharge over the long-term, and to collaborate with the division of urology to spread the algorithm to other providers caring for this population. Future iterations will also include a recommendation for pelvic ultrasound for the subgroup of female adolescent patients with CSK at the 15-year time point to identify possible associated genital anomalies before transition of care to primary care. We also plan to work on a transition of care plan for patients to increase provider and patient comfort with eliminating ultrasound appointments and increase long-term adherence to, and sustainability of, the algorithm.
Acknowledgments
We acknowledge and thank Mickey Jawa for his mentorship and technical guidance throughout the duration of this project, particularly with data analysis and interpretation. We also acknowledge and thank Monica Piekut, RN, and Dr Amrit Kirpalani for their input on the diagnostic assessments and development of the algorithm.
Ms Jawa designed the study, designed data collection instruments, collected data, conducted the analyses, drafted the initial manuscript, and reviewed and revised the manuscript; Dr Matsuda-Abedini conceptualized the study, supervised data collection and analysis, and reviewed and revised the manuscript for intellectual content; Drs Rosenblum and Radhakrishnan supervised data analysis and reviewed and revised the manuscript; Drs Pearl and Levin critically reviewed the manuscript for intellectual content; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
FUNDING: No external funding.
Comments