Table 3.

Summary of challenges encountered in implementing the Clinical Data Management (CDM) framework for CIN and solutions developed (normal text) or proposed (italicized text).

Challenge	Details of Challenge	Long-term Solution
Resource limitations	Internet connection challenges, staffing challenges, insufficient computers, power outages, software licensing costs	1. Hire a second clerk to each CIN hospital to support the network 2. Provide internet modem to all clerks to use to send data 3. Provide a computer to each hospital for data capture 4. Use open-source /freeware tools to support data capture—i.e., machine running on Ubuntu, REDCap for data capture, R for analysis 5. Use UPS units to mitigate effects of power outages
Adoption of data codification standards	Use standard nomenclature to code variables used in pediatric data at point of data transfer from paper to electronic form	1. Use lookup lists for: – Treatment fields to capture store generic drug name when brand name entered – Diagnosis fields to capture values as their ICD-10 equivalent 2. Implement SNOMED-CT equivalent for all CIN variables and make the code book available when sharing data 3. Collaboration with Ministry of Health and Kenya Paediatric Association, hospital clinical staff to adopt a standard clinical data model for pediatric patients tied to SNOMED-CT / ICD-10
Data synchronization	Automate data consolidation	1. Update synchronization module to allow data export in batches 2. Set the maximum limit of data transfer to observations not older than 30 days
Data quality control	Ensuring good quality data is being captured	1. Create standard operating procedures for data collection 2. Use cleaning scripts to validate accuracy of data entered at the hospital level 3. Conduct bimonthly data quality assurance exercise, using a sample of previously entered records 4. Generate daily data quality reports and use them to go through any errors with data clerk over a telephone call
Data collection tool creation and update	Integrate CIN data collection with routine job aides and work flows, allow for updates	1. Design data abstraction of paper records to REDCap tool to reflect CIN hospital workflow and to ensure data elements needed for DHIS reports included 2. Data entry from paper records to REDCap tool to happen after patient discharge/death 3. Update REDCap tool the same time as when conducting bimonthly data quality assurance exercise
DHIS Reports for Ministry of Health	Generate and submit DHIS reports for each CIN hospitals directly to Ministry of Health	1. Create Script to Generate DHIS reports based on pre-specified cohort groups for pediatric care 2. Engage the Ministry of Health to allow automated submission of DHIS reports to the national HIS through the web API —this is still ongoing

API = application programming interface