Protecting Infants Remotely by SMS (PRISMS)

Countries in resource limited settings contribute over 90% of the global burden of neonatal mortality. Neonatal mortality (<28 days after birth) is largely preventable through effective preventive measures to improve neonatal survival. These newborn survival measures include widely available interventions such as: 1) timely administration of antibiotics, 2) Kangaroo mother care, and 3) alternate methods of feeding such as expressed breast milk.  Despite the availability of these effective interventions, neonatal mortality rate remains unacceptably high at 27 per 1000 live births in Uganda. In resource rich settings, neonatal mortality rate is low and neonatal care is a highly specialised discipline. In developed country settings, decisions regarding sick newborn care are made by highly qualified pediatricians, specialist neonatal nurses and or neonatologists. In resource limited settings, the bulk of sick newborn care management decisions are made by medical officers, nurses, and or midwives who have not received specialised training. These staff cadre typically have inadequate training to make management decisions for sick newborn babies. Therefore, we developed and tested the impact a health worker-targeted newborn care tool called “Protecting Remote Infants by SMS (PRISMS)”. We assessed it's impact on 1) reduction of duration of hospitalisation and 2) reduction of in-facility neonatal mortality with significant results. PRISMS leverages routine clinical assessment findings to provide the non-specialist health workers with timely clinical newborn management guidelines to care for sick newborns. 

Protecting Infants Remotely by SMS (PRISMS) is a mobile platform that provides clinical decision support for non-specialist health workers in marginalised populations in Uganda and Africa.  PRISMS uses routine assessment findings collected during review of admitted babies to generate care plans for health workers. These collected data points from the assessment are locally stored on the mobile device used. This enables the health worker to work without the need for telephone network. However, in presence of internet, the devices automatically synchronise all the data with a remote server aggregated according to the facility. The server has enabled us to; i) Generate quality improvement data to guide targeted quality improvement meetings, ii) Build model for early prediction of referral and mortality, and iii) Enable health care planning and surveillance. PRISMS has been currently used on more than 100,000 babies  from 66 facilities. PRISMS has potential for rapid nation-wide deployment. Our initial study in Uganda showed that: 1) PRISMS server can process over 60,000 simultaneous requests with no error and time lag of no more than 5 seconds, 2) all health workers agreed or strongly agreed that PRISMS reminded them of aspects of care they had missed to offer, and 3) there was statistically significant and high agreement between Pediatrician and PRISMS suggested management. Further learning from this and the ongoing RCT informed the current version of PRISMS that is not internet and telephone network dependent, and enabled to provides longitudinal patient care and follow-up. 

In a country like Uganda with a total population of 41 million people with a national crude birth rate of 42.6 per 1000 people (2015), there are about 1.7 million births per year. With neonatal mortality rate accounting for 41% of all under 5 years deaths, PRISMS has the potential to save a significant number newborn lives with large scale impact if implementation is carried to national level. We currently recommend PRISMS use at admission, ward rounds and during pre-discharge newborn assessments at maternity wards.  All newborn care health providers stand to benefit from PRISMS. Uganda has a total of 1,529 birthing facilities (HC III, HC IV and Hospitals) with no specialist doctors. With a conservative estimate of 5 newborn care providers per care center (Midwives, clinical officers, medical officer and or nurses), a total of 7,645 health providers stand to be empowered by PRISMS. This conservative estimate however excludes the private midwives, and other private clinic care providers across Uganda. It also excludes preservice medical and nursing students.

Adoption of PRISMS by health care providers in Uganda is expected to provide uniform management of sick and at-risk newborn babies with potential to harmonise standard of care and save lives. We know that PRISMS to enhance rapid and timely appropriate medical interventions that will lead to overall improvement in the quality of medical care given to sick newborns. We anticipate that access to information on appropriate care will consolidate knowledge renewal and retention by health workers, build confidence for newborn care and, with time, lead to sustained improvements in the quality of newborn care delivered. It is also an adaptable platform capable of incorporating and reinforcing alterations in accepted care pathways. Ultimately, PRISMS has the potential to significantly improve the overall survival of newborn babies in Uganda and in other countries with similar challenges for newborn care. 

The innovation team is based at Mbarara University of Science and Technology and also implements community health interventions. This is the team that developed this system to guide health workers access timely support in management of newborns. The team has regular interaction with health workers working on maternal child health issues especially in Government supported health facilities in Uganda. The innovation team presents updates regularly to stakeholders and obtains feedback to improve the system.

Like one critical piece of advice we received was during the Innovations Market place at the Regional Interagency meeting on Neonatal Health for East and Southern Africa Region organised by UNICEF. This was about the system architecture with a request to achieve phone based patient specificity, attain automatic growth rate and feeds calculation. The system architecture and implementation of most of the core modules was quickly revised with all requested features implemented, tested and critically appraised. This revision has since improved PRISMS performance reflected in PRISMS version 5. This is a good example of how people closest to the problem have the potential to iterate the solution with ease. 

Additionally, the innovation team has been involved in the design and development of solutions for marginalised populations for more than 15 years. 

For the last 5 years, we have seen the impact of PRISMS on reduction of duration of hospitalisation and in-facility mortality in hospitals in marginalised populations because of decision support. We also note that most health workers in marginalised populations lack sufficient training to make clinical decisions. Lastly, we did a deep dive into the existing Electronic Medical Records (EMR) existing in Uganda, Burundi and Tanzania and we noted that most of these don't have decision making support but are rather used as data collection tools to guide programmatic planning.

We have leveraged our experience from PRISMS development to start designing an EMR system with Brains (one that has the potential to guide decision support). We shall transverse all hospital departments and build a solution that non-specialist health workers can use to guide their decisions in their day to day activities. With additional support, we hope to have the solution completed in the shortest time possible and deployed in additional health facilities in Uganda and Tanzania to contribution to saving more lives.

Neonatal mortality is still the biggest contributor of the under five deaths. In most of African health setting structures, there are no Pediatricians at the lower level health facilities which are the first point of contact for community before referrals to highly specialized  levels is done. The cases are mostly handled by nurses and midwives who do not have the specialization of management like a Pediatrician would. PRISMS, has been designed to empower these lower cadre health workers to manage sick newborns like a Pediatrician would, "Call it a Pediatrician in your pocket."  From the validation trials done, PRISMS has decreased referrals to higher levels of care , empowered lower health cadres to effectively manage sick newborns and most importantly reduced  deaths in facilities where it has been deployed.  PRISMS is therefore a "pediatrician device" to ease access to highly specialized care to newborns at low cost and save lives.

PRISMS, has been validated and is effective in the management of sick newborns. In the next one year, we would love to see a bigger uptake of PRISMS up to about 50% coverage of health centre IV and District Hospitals that do not have Pediatricians. In the next 5 years, we would like scale up of up to 90% coverage of all the lower level health facilities that conduct deliveries. The one year impact can be made possible with this funding if provided for this innovation. The 5 year impact can be best  achieved through partnerships with Ministry of Health and other health implementing agencies in reducing neonatal mortalities. Working with Ugandan Ministry of Health will enable us scale up this innovation to World Health Organization in the next 5 years.

We look forward to contributing towards a number of indicators that include;

1. Increased number of children attended to by a skilled health worker

2. Reduced neonatal mortality in Uganda and the World

3. Reduced child mortality rate

Our theory of change states that;

Increased access and utilization of PRISMS apple among health workers, will improve application of technical skills and utilization of expert pediatrician  knowledge and skills, this will greatly improve neonatal treatment outcomes and increased chances of child survival.

Protecting Infants Remotely by SMS (PRISMS) is a Short Message Service (SMS) that empowers frontline health workers.  PRISMS helps health workers to effectively provide newborn care clinical management through provision of instant clinical care decisions based on routine assessment findings.  Frontline health workers input routine newborn clinical assessment findings into PRISMS and send these findings by SMS to a remote automated server. The server processes these patient-specific clinical-assessment-findings and provides management suggestions by SMS. PRISMS’ SMS response time is 8 seconds as determined in a recent multicenter trial conducted in four regional referral hospitals in Uganda. PRISMS has potential for rapid nation-wide deployment. Our initial study in Uganda showed that: 1) PRISMS server can process over 10,000 simultaneous requests with no error and no time lag of more than 8 seconds, 2) all health workers agreed or strongly agreed that PRISMS reminded them of aspects of care they had missed to offer, and 3) there was statistically significant and high agreement between Pediatrician and PRISMS suggested management.

Management suggestions are processed from the PRISMS server based on all form inputs and sent to the front line health worker within an average of 8 seconds as a complete and holistic clinical management plan. Clinical assessment findings input into PRISMS and sent to the server are stored at the server end together with its corresponding management suggestion sent to clinician. This instant and continuous data collection process is critical for quality improvement, burden of disease surveillance and for objective newborn service planning at facility, district, regional and national level. Health workers motivation is the quality of pediatric support availed through the apple to save lives of newborns.

E-Wall Technologies works with all people of all tribes, race and cultural background. We do not discriminate any staff or beneficiary based on any traits but rather embrace diversity and inclusiveness as opportunities for strength of our work and ideas. Employment is based purely on competence and capacity of the member and all sex is welcome to work with us.