Written By: Allan Wasega (Data Analyst, nLine), Margaret Odero (Data Analyst, nLine) and Jordan Fast (Solar for Health Energy and Climate Specialist, United Nations Development Program (UNDP))
Reliable electricity as a pillar for healthcare delivery
Reliable and quality electricity is one of the key ingredients in the provision of quality healthcare. Various healthcare functions, ranging from storage of vaccines to the operation of medical equipment such as ventilators and defibrillators, are likely to fail without a stable and dependable electricity supply - with dire consequences to patients. In-healthcare facility mortality in Ghana, for example, is estimated to rise by 43% for each day power is out for over two hours. Alarmingly, only 50% of hospitals in sub-Saharan Africa have reliable electricity access, while 15% lack access completely. Given the critical role healthcare plays in ensuring the wellbeing of people and, by extension, the economies of nations, energizing healthcare has been a major focus of governments and other stakeholders. Sustainable Energy for All (SEforALL), for example, has made “universal healthcare electrification by 2030” a key part of its impact areas, and has been working with various partners to electrify healthcare facilities in Sierra Leone. Among the measurements used to assess the success of these electrification efforts are the number of facilities brought online, solar system (and other generation sources) uptime, and the frequency and duration of outages at an establishment. In this post, we describe findings from measuring both hospital- and room-level KPIs, uncovering that different rooms or departments within healthcare facilities experience varying PQR outcomes. In doing so, we highlight the importance of considering room-level KPIs, especially in critical rooms, alongside hospital-level metrics to more accurately measure healthcare electrification outcomes.
nLine's contribution in reporting healthcare electrification metrics: Should we explore multi-level aggregation?
Healthcare electrification projects often encounter key data challenges. A Sierra Leone market assessment report by SEforALL revealed that data on healthcare electrification is often unavailable, and where it exists, it typically does not includes reliability indicators and has granularity challenges. Another common challenge is the absence of clear benchmarks for defining minimum acceptable electricity access at the hospital level, which complicates evaluation efforts. During their healthcare electrification assessment in Sierra Leone, for example, the former Crown Agents faced difficulty determining whether the PQR improvements achieved through solar installations were adequate. Without predefined benchmarks, it was challenging to assess whether the PQR levels met the needs of the healthcare facilities and to confidently evaluate the success of the electrification efforts. Since 2021, nLine has been instrumental in filling these data gaps by providing comprehensive granular power quality and reliability (PQR) measurements to aid in monitoring, evaluation, and learning (MEL) efforts using its GridWatch technology. Building on these efforts, nLine is supporting micro and mini-grid healthcare electrification assessments in more than 63 healthcare facilities across Sierra Leone for SEforALL and defunct Crown Agents, as well as in the DRC, Rwanda, Kenya and South Sudan — Figure 1.
This article uses data from nLine’s GridWatch sensor deployment in Sierra Leone. This deployment was done across multiple departments in 17 healthcare facilities that are part of SEforALL’s Sierra Leone Healthcare Electrification Project. Each monitored room/department in the healthcare facilities was fitted with one GridWatch sensor, and each healthcare facility had at least 3 in total. These sensors served two primary purposes: establishing baseline PQR metrics to inform the design and procurement of solar PV and battery storage systems, and measuring how well the newly installed renewable energy systems addressed existing PQR challenges in these facilities. The collected data enabled SEforALL’s former implementing project partner, Crown Agents, to quantitatively evaluate how well the newly installed renewable energy solutions address existing power quality and reliability challenges in these facilities.
The data also revealed to us some variations in electrification benefits across departments within the same healthcare facility. In particular, some departments experienced significantly greater PQR challenges than suggested by facility-wide averages. This raises an essential question: While facility-level improvements are evident, how effectively do these benefits extend to individual departments, especially critical care areas? Understanding this granular impact is vital to linking electrification efforts to specific healthcare outcomes like maternal mortality and morbidity, and could also be key in defining what constitutes “minimum acceptable” electricity access at the hospital level.
To maximize the benefits of healthcare electrification, we also propose that it's crucial to prioritize "critical" rooms, given that they host equipment and services most dependent on reliable electricity and are also the rooms that provide critical, lifesaving services.
What are “critical” rooms in a healthcare facility?
Early pioneers of modern healthcare, such as Florence Nightingale, recognized the importance of organization in medical care, including the need to group the sickest patients into designated areas where they could be closely monitored and receive specialized attention. Building on such foundational principles, today’s healthcare facilities prioritize certain areas — such as intensive care units (ICUs), operating rooms (ORs), emergency rooms (ERs), and neonatal intensive care units (NICUs) — as "critical" due to the essential roles they play in healthcare provision. These rooms require uninterrupted power supply for life-saving care and patient survival.
ICUs, for instance, house patients in critical condition who require life-support machines, ventilators, and real-time monitoring. Power disruptions in these rooms can have severe consequences, as even brief interruptions in equipment functionality can directly impact patient outcomes and even lead to death. Similarly, operating rooms depend on stable power to support complex surgical equipment, including anesthesia machines, electrocardiograms (ECGs), and surgical lighting. Any power interruption mid-surgery can disrupt critical procedures, putting patients at heightened risk and necessitating rapid re-stabilization efforts. NICUs further underscore the importance of reliable power, as critically ill and premature newborns depend on incubators, heart monitors, and ventilators for survival. Studies indicate that any power instability in NICUs correlates with increased health risks due to equipment malfunction, emphasizing the critical need for backup power solutions in facilities that rely on off-grid systems or with no access to the grid.
The case for room-level healthcare electrification metrics
This section provides illustrative cases from four healthcare facilities where solar PV and battery storage system installations were done under SEforALL's Sierra Leone Healthcare Electrification Project. During this project, the new solar systems went live around the beginning of January 2024. Through these case studies, we demonstrate the importance of adopting critical room- and facility-level PQR metrics to comprehensively assess the impact of healthcare electrification efforts. We examine two criteria: the duration of undervoltage episodes and the average duration of power outages.
1. Average daily duration of undervoltage
Aggregating data by averaging can mask important variations in power quality and reliability across different parts of a healthcare facility. For example, undervoltage, defined as periods when supply voltage drops below 207V (90% of Sierra Leone's nominal 230V), demonstrates markedly different patterns when examined at facility-wide versus room-specific levels. Figures 3a and 3b illustrate the difference in undervoltage hours recorded at the hospital - versus room-level at Kambia Government Hospital.
The mean daily undervoltage hours at the hospital-level range from less than 1 hour to 7 hours per month (Figure 3a). However, at the room-level, these values vary significantly, ranging from 0 to over 20 hours per day, as in the case of the pediatric resuscitation room which experienced prolonged undervoltage durations of up to 22 hours per day in March (Figure 3b), whereas the hospital-level reading was 4.25 hours per day (3a). Experts recommend that emergency medical personnel in pediatric resuscitation rooms should have access to certain essential equipment, including automated external defibrillators (AEDs), pulse oximeters, and CO2 monitors. These devices require stable electricity to function optimally; otherwise, they can return unreliable readings, suffer slower response times, or fail to function completely. These challenges heighten the risk of poor outcomes for pediatric patients who need rapid and accurate interventions during critical emergencies.
Figure 3c, a time series representation of the two-minute voltage data for March 2024, further shows the voltage values from the different rooms at Kambia Hospital. The voltage values at the paediatric resuscitation room (in red) are predominantly below the nominal band of +/-230V for the entire month.
In seeking to understand the cause of this consistent undervoltage at the paediatric resuscitation room in Kambia hospital, we sought insights from hospital staff. We learned that this critical room operates on a standalone solar system, separate from the newly installed solar infrastructure serving the rest of the hospital. This lack of integration has resulted in uneven PQR across different rooms in the same hospital. As a result, critical rooms like the pediatric resuscitation room remain disproportionately affected, despite the implementation of progressive electrification initiatives. This highlights two issues. First, it is important to evaluate the success of healthcare electrification interventions at a granular level and not assuming that the addition of a new power source automatically improves access to quality electricity across the facility. Second, while reporting healthcare electrification outcomes using the undervoltage hours metric at the hospital-level, nuances such as extremely poor voltage quality in critical rooms are missed.
2. Average power outage duration
nLine’s outage detection algorithm works by clustering power-off reports from sensors that occur at around the same time within the same geographical location. In health facilities in Sierra Leone, at least three nLine sensors are installed at each facility. A hospital-level outage is confirmed when at least two sensors report a power-off event simultaneously. These clustered outages are used to compute average outage duration (SAIDI) at the hospital-level.
To estimate daily uptime (percent of time in the day when power is available), we use:
Figure 4 shows the hospital-wide uptime from four off-grid hospitals between January - September, 2024.
The uptime data indicate mixed results. In Masanga and Kabala, for example, system uptimes of above 60% are consistently recorded across all months, suggesting that power is available for at least 18 days in a 30-day month (or, conversely, unavailable for 12 days - 288 hours) in a 30-day month. However, sensor measurements at the room-level at these two hospitals reveal slightly longer monthly outage durations in certain areas of the hospitals, including critical rooms (See Figures 5a and 5b). In April, for instance, the Emergency Room at Masanga suffered outages lasting 245.67 minutes (10.2 days) in total (Figure 5a). In contrast, the hospital registered a system uptime of 84% (Figure 4), representing a total SAIDI of 4.8 days. In another case, the Paediatric Pharmacy at Kabala registered outages of 342.25 minutes (14.26 days) in February (Figure 5b), against a hospital-wide SAIDI of 6.9 days for the month.
These monthly room-level measurements of SAIDI at Kabala Government Hospital reveal critical gaps between overall system performance and room-specific measurements, underscoring the importance of granular assessments in evaluating healthcare electrification outcomes. While hospital-wide solar system uptime indicates excellent energy availability, certain critical rooms face significant interruptions that threaten essential healthcare services. For instance, the serology room reported outages of over 200 hours (8.33 days) in June and July, which can jeopardize diagnostic processes where blood samples must be tested within six hours to remain viable. Similarly, the pediatric pharmacy experienced over 400 hours (16.67 days) of outages in January. Conversations with hospital staff revealed that there were wiring issues which may be resulting in such prolonged outages in the pediatric pharmacy, highlighting how underlying PQR challenges can remain obscured by system-wide metrics. These findings emphasize the necessity of room-level measurements to identify and address PQR disparities, ensuring reliable power for all critical hospital functions.
Thus, while aggregated metrics offer a high-level perspective on electrification performance, disaggregated room-level data provide crucial context for evaluating impact of electrification on critical care spaces where voltage quality and power reliability directly affects life-saving interventions.
Conclusion
Electrifying health facilities across Africa holds immense potential to enhance patient care and health outcomes. Presently, however, the impact of these interventions is often assessed through generalized, facility-wide metrics. These aggregated indicators, while useful for broad assessments, can obscure critical vulnerabilities in the areas where reliable power is most essential — in critical rooms like ICUs, operating rooms, NICUs, and emergency wards. nLine's observations in healthcare facilities across Sierra Leone reveal that despite improved system-wide PQR measures such as uptimes, room-level data sometimes tell a different story. Critical rooms may experience frequent power disruptions or voltage instability, compromising the functionality of life-saving equipment and jeopardizing patient well-being. This discrepancy underscores the importance of developing and adopting room-level PQR metrics to provide a more accurate picture of healthcare electrification outcomes.
Future interventions should incorporate room-level metrics and be guided by lessons learnt from past deployment processes. For instance, in monitoring room-level KPIs, we should employ a deployment methodology that allows for installing multiple sensors per room without using up all power sockets. Granular, room-level data could also inform more robust KPIs for evaluating healthcare electrification success. For instance, critical rooms could have specific availability thresholds, ensuring that lifesaving equipment and services remain operational. Defining such thresholds — for example, setting minimum electricity availability KPIs for critical rooms — could provide clearer targets for electrification projects and better align energy access with healthcare outcomes. Finally, this blog post only uses data measured by nLine’s Gridwatch system in the analysis. We can combine the PQR data from nLine sensors with those from other sources used in these projects such as consumption and load data, and solar system uptimes to enrich the analysis and insights and perform more robust impact analysis on these healthcare electrification projects.
As we continue to monitor and analyze data from healthcare facilities, nLine is committed to advocating for methodologies and solutions that prioritize precision and localized insights in addressing healthcare electrification challenges.
References
[1] Apenteng BA, Opoku ST, Ansong D, Akowuah EA, Afriyie-Gyawu E. The effect of power outages on in-facility mortality in healthcare facilities: Evidence from Ghana.
Global Public Health. 2016;13(5):545-555. doi:https://doi.org/10.1080/17441692.2016.1217031
[2] Electricity in health-care Facilities. Published August 31, 2023. https://www.who.int/news-room/fact-sheets/detail/electricity-in-health-care-facilities
[3] Sustainable Energy for All. "Sierra Leone Healthcare Electrification Project." Accessed November 22, 2024. https://www.seforall.org/programmes/powering-healthcare-hub/sierra-leone-healthcare-electrification-project.
[4] Sustainable Energy for All (SEforALL) on LinkedIn. Published October 8, 2024. https://www.linkedin.com/posts/seforall_sierraleone-sustainableenergy-poweringhealthcare-activity-7249366354972033026-GUzf?utm_source=share&utm_medium=member_desktop
[5] Esper AM, Arabi YM, Cecconi M, et al. Systematized and efficient: organization of critical care in the future. Critical Care. 2022;26(1). doi:https://doi.org/10.1186/s13054-022-04244-1
[6] Molefi Sompane. Free State health is investigating patient death after generator runs out of diesel during power outage. Health-e News. Published July 15, 2024. https://health-e.org.za/2024/07/15/free-state-health-is-investigating-patient-death-after-generator-runs-out-of-diesel-during-power-outage/
[7] Messersmith LJ, Ladha A, Kolhe C, et al. Poor power quality is a major barrier to providing optimal care in special neonatal care units (SNCU) in Central India. Gates Open Res. 2022;6:58. Published 2022 Apr 27. doi:10.12688/gatesopenres.13479.1
[8] Kaur E, Heys M, Crehan C, et al. Persistent barriers to achieving quality neonatal care in low-resource settings: perspectives from a unique panel of frontline neonatal health experts. Journal of Global Health Reports. 2023;7. doi:https://doi.org/10.29392/001c.72089
[9] López-Herce Cid J, Rodríguez Núñez A, Carrillo Álvarez Á, Zeballos Sarrato G, Martínez Fernández-Llamazares C, Calvo Macías C. Materials for the paediatric resuscitation trolley or backpack: Expert recommendations. Anales de Pediatría (English Edition). 2018;88(3):173.e1-173.e7. doi:https://doi.org/10.1016/j.anpede.2017.05.002