https://orcid.org/0009-0007-9130-8754
Figures
Abstract
Infection prevention and control is essential for high-quality health care. Chlorine is a widely used disinfectant in health care facilities; however, inconsistent availability and quality pose challenges for health staff. On-site chlorine production offers a potential solution. This study evaluates the reliability, change in chlorine availability and quality, and total cost of ownership of the Aqua Research STREAM Disinfectant Generator in primary health care facilities in Ghana. Our cross-sectional study evaluated 18 STREAM devices across 12 primary health care facilities in Ahafo, Central, and Volta Regions. A mean time between failures approach was used to benchmark device reliability. Historical commercial chlorine stock records were compared with STREAM chlorine production volumes. Chlorine quality assessments compared free residual chlorine levels in commercial and STEAM chlorine samples. A 5-year cost model estimated the total cost of STREAM ownership and per-liter cost savings potential. STREAM units remained functional for 94.8% of the study; 67% (n = 12) functioned without failure. Six devices averaged 60.8 days before their first failure (range 16–96 days). No device recorded more than one component failure. STREAM devices eliminated chlorine stock outs, previously occurring 37 days per year on average. In all, 83% of health facilities were found to have degraded commercial chlorine, whereas 100% of STREAM samples met the device’s target concentration (0.5% ± 0.1% mg/L). The 5-year total cost of ownership ranged from $3,700 to $7,900. District hospitals saw a 17% reduction in annual chlorine supply costs with STREAM units, while health centers experienced an increase of 48%. The STREAM shows promise for improving infection prevention and control practices in Ghana’s health system. Study results informed the Ghana Health Service’s decision to purchase 400 STREAM devices (June 2024). Future research should explore operational models that expand chlorine availability and lower chlorine costs in health facilities.
Citation: Drolet A, Cofie P, Aburam K, Yaa-Dapaah Boakye E, Hsu S, Orndorff C, et al. (2026) Improving infection prevention and control in Ghana primary health care facilities: Evaluation of the STREAM disinfectant generator. PLOS Water 5(1): e0000383. https://doi.org/10.1371/journal.pwat.0000383
Editor: D. Daniel, Gadjah Mada University Faculty of Medicine, Public Health, and Nursing: Universitas Gadjah Mada Fakultas Kedokteran Kesehatan Masyarakat dan Keperawatan, INDONESIA
Received: April 17, 2025; Accepted: January 5, 2026; Published: January 30, 2026
Copyright: © 2026 Drolet et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All relevant data are within the paper and the Supporting information files.
Funding: This work was funded by the Conrad N. Hilton Foundation under grant #28708 (AD). The funding agency had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
1. Introduction
Health care–associated infections (HAIs) affect hundreds of millions of individuals each year [1,2]. Patients in low-resource countries face a disproportionately greater risk of HAIs—3–20 times higher—than patients in similar settings in high-income countries [3]. In Ghana, an estimated 303,000 HAIs, half of which were predicted to be antimicrobial resistant, led to 31,400 deaths in 2022 [4]. Infection prevention and control (IPC) programs are highly effective in reducing the spread of pathogens and infections in health care settings and critical to the delivery of high-quality, safe health care services [5]. Pooled analyses from systematic reviews indicate IPC programs—which include the use of disinfectant such as bleach for environmental hygiene practices—can reduce HAI rates by 35% to 70% [6–8]. Hand hygiene and environmental hygiene interventions have been identified as leading IPC cost-saving measures for health systems. Implementation and adherence to hand hygiene programs is estimated to reduce health expenditures by US$16.50 for every $1.00 invested. When combined with environmental hygiene improvements, the risk of mortality from antimicrobial-resistant pathogens is expected to decrease by up to 49% [9].
The Ghana Health Service (GHS) has focused on enhancing IPC through various strategies and policies. The 5-year (2024–2028) National Infection Prevention and Control Strategy introduced updated guidance to strengthen IPC capacity, auditing, and HAI surveillance systems across the country [10]. Additionally, the National Healthcare Quality Strategy (2017–2021) emphasized the need for consistent availability of IPC supplies, equipment, and essential water, sanitation, and hygiene (WASH) infrastructure in health care facilities [2]. Since its launch in 2016, the national IPC/WASH task force has played a key role in institutionalizing professional training and capacity-building of health workers; however, challenges remain. An assessment of IPC readiness across 56 health facilities in Ghana identified significant gaps, including the absence of well-defined IPC objectives, insufficient budgets, irregular mandatory training, and limited access to essential IPC materials [11].
Chlorine is a commonly used and widely recommended disinfectant for cleaning health care environments; yet, many health facilities worldwide struggle to maintain adequate chlorine supplies due to challenges such as inefficient supply chains, complex procurement procedures, and limited financial resources [12,13]. A global review covering 78 low- and middle-income countries estimated that 36.4% of health care facilities lacked chlorine for disinfection [14]. Additionally, a facility assessment conducted in November 2020 on chlorine stock records from eight health care facilities in Ghana’s Eastern Region—including district hospitals and health centers—revealed that facilities experienced an average of 44.8 days per year without chlorine [15]. Beyond availability issues, research has highlighted concerns regarding the use of degraded disinfectants, which may reduce effectiveness, contribute to antimicrobial resistance, and fail to inactivate pathogens responsible for HAIs in Ghanaian health care settings [16,17].
Electrolytic chlorine generators provide a promising solution for health care facilities facing chlorine shortages and inconsistent product quality. While various options are available, many on-site chlorine generators are unsuitable for low- and middle-income countries due to high capital costs, reliance on proprietary components, need for continuous electricity, and the requirement for skilled personnel to oversee operation and maintenance. Electrolytic chlorine generators designed for low-resource settings, such as the STREAM™ Disinfectant Generator (Aqua Research, New Mexico, USA), offer a practical solution for health care staff, enabling on-demand production of high-quality chlorine using simple inputs (salt, water, and electricity). Whereas batch-based chlorine generators require long periods of consistent energy to complete a chlorine production cycle—including the use of reagents and specific testing equipment if electricity is cut during the production cycle—the STREAM’s automated and flow-through production system ensures that the chlorine generated by the STREAM will consistently have a concentration of 5,000 mg/L and can be used immediately. In Ghana, chlorine generators have been installed in at least 25 health care facilities across seven regions (Eastern, Northern, NorthEast, Oti, Savannah, Upper East, Upper West), and further national expansion is being led by the GHS.
PATH and the GHS Institutional Care Division sought to address a critical knowledge gap regarding the feasibility, reliability, and effect of on-site chlorine production to address chlorine stock outs, inconsistent quality of chlorine, and potential improvements to IPC practices in Ghanian health facilities. To this end, PATH collaborated with the GHS Institutional Care Division to evaluate the STREAM device in 12 health care facilities in Ahafo, Central, and Volta Regions. This electrolytic chlorine generator, designed for low-resource settings, produces a steady 0.5% mg/L (5,000 parts per million) chlorine solution. The primary objective was to assess the reliability of the device in health care settings. Secondary objectives included assessing the device’s impact on chlorine availability and quality and determining the overall cost of ownership of the STREAM.
2. Materials and methods
Ethics statement
Ethics approvals were obtained from the WCG Institutional Review Board (1933773–1, September 2, 2022) and the GHS Ethics Review Committee (GHS-ERC 020/08/22, March 9, 2023). Additional information regarding the ethical, cultural, and scientific considerations specific to inclusivity in global research is included in the Supporting Information (S6 Checklist).
Health facility and staff recruitment
Our study incorporated various types of health care facilities to assess chlorine demand across different levels and facility types and to evaluate how effectively the STREAM devices met their chlorine needs. A convenience sample of 18 STREAM units were deployed and evaluated across 12 health facilities—six district hospitals, four health centers, and two polyclinics—located in Ahafo, Central, and Volta Regions (Table 1).
The GHS selected the regions, districts, and health facilities for this study with the intent of generating a broad geographical and health facility–level sample for the purpose of comparison. Monthly volume usage rates for disinfection were between 400 and 800 liters of 0.5% mg/L or 5,000 parts per million chlorine solution (roughly 20–40 liters per day over a 5-day work week). Study sites were required to have a reliable electricity supply (mains or generator) and access to a water source (mains, borehole, etc.) on the premises. All health facilities had updated commercial chlorine stock records, with data showing volume received and distributed within the health facility. Hospital management at each location gave approval to conduct the study and continue STREAM device operation following the conclusion of the study.
STREAM placement within each health facility was determined by the hospital administrator and health staff. Criteria used for selecting the installation location included centralized, secure, and easily cleaned in case of occasional or accidental chlorine or water spillage; proximity to a water source; and reliable and consistent electricity access.
Prior to recruitment of the 12 participating health facilities, study staff presented each hospital administrator with a health facility participant information sheet that described the study objectives, risks and benefits, participation and withdrawal, and confidentiality procedures. Signed consent was obtained from all 12 facility administrators. Given no personal information, opinions, or biospecimens were collected from individuals, informed consent was not collected from STREAM operators or users.
User training sessions were conducted in each participating health facility with three primary STREAM operators and health care personnel responsible for using STREAM chlorine for disinfection and environmental cleaning purposes. These sessions, approximately 60 minutes in duration, covered essential topics such as device assembly, operation, troubleshooting, cleaning, and maintenance. To support ongoing use, training materials and operational guides were provided at each site for reference.
Study design
Our cross-sectional study evaluated the reliability, impact on chlorine availability and quality, and overall cost of ownership of the STREAM device in primary health care facilities in Ghana for 6 months (June 2023 through November 2023). The reliability of the STREAM device was evaluated using a mean time between failures approach [18]. This approach measured the time in days from recorded device failure to device repair in order to calculate the percentage of the study in which devices were operational. The assessment considered the total number of failures, and the average number of failures reported during the evaluation period, along with the average operational time between failures and the proportion of time the units remained functional. These metrics were derived from the operation run time data automatically gathered by the STREAM devices and daily STREAM chlorine production values recorded by device operators on STREAM monitoring forms. Failures were identified and categorized separately from errors in this analysis. Failures were defined as the malfunctioning of one or more components that caused the entire STREAM unit to be nonoperational for more than 24 hours. These components were either repaired or replaced. In contrast, errors referred to issues arising from user actions, context, or mechanical problems that could be diagnosed and fixed by users within 24 hours with or without assistance from PATH staff or the manufacturer. At each study site, users were provided with a STREAM monitoring form to document errors and failures as they occurred. The failure period was marked by the recorded failure date, and the period ended once the device was functional again following repairs. The duration and frequency of failure periods were then calculated. The operational status of the devices was tracked through phone calls and in-person visits by the research team, which also recorded specific device errors, component issues, and replacement parts installed for each unit.
STREAM functionality and error data, including total operational hours, production volumes, and error and failure occurrences and dates, were cleaned and compiled in a Microsoft Excel database. The data were used to calculate the percentage of time the devices were functional during the evaluation period, the average time from installation to device failure, and the average time to repair reported failures. Users who reported device failures were able to identify the causes (e.g., site-specific fluctuations in electricity supply and user error) and faulty components, which allowed for calculating components with the highest failure rates during the evaluation period. Component failures were defined as those requiring replacement or significant corrective action by PATH staff or regional biomedical engineers. Minor issues that could be addressed by health facility staff were excluded.
Commercial chlorine inventory logs (stock cards) from each facility were examined and recorded at the outset, covering data and images from a 3- to a 12-month period from June 2022 to June 2023 (prior to the evaluation). Historical chlorine stock–level data were used to generate monthly chlorine demand levels, creating a benchmark for comparison with the actual monthly volumes of chlorine produced by the STREAM device. Commercial chlorine stock card data were also reviewed, to determine the frequency and duration of commercial chlorine stock outs in study facilities (i.e., instances when chlorine stocks were depleted for more than 24 hours). Stock outs of STREAM-generated chlorine were defined as any period longer than 24 hours during which the STREAM unit was not operational. Health facilities were encouraged to rely solely on the STREAM for chlorine production throughout the study and use commercial chlorine only if the STREAM device failed or if there was a need for additional chlorine.
Chlorine stock cards were reviewed at baseline, monthly during monitoring visits if any commercial chlorine was received, and at endline by PATH study staff and accompanying GHS Institutional Care Division counterparts to determine if additional commercial chlorine was procured and distributed along with the chlorine generated by the STREAM device. Chlorine volumes were tracked using specific forms: STREAM monitoring forms for chlorine generated by STREAM devices and commercial chlorine stock monitoring forms for any supplementary commercial chlorine received. Daily, primary STREAM operators manually recorded the volume of chlorine produced (in liters) and the wards in which the STREAM chlorine was distributed. These forms were collected by the research team during monitoring visits. Furthermore, the STREAM device’s internal data logger automatically tracked and stored total operational hours, which the team retrieved during monthly site visits and used to generate quantitative measures of total chlorine production.
The quality of both commercial chlorine and STREAM-generated chlorine was assessed using the Hach DR300 pocket colorimeter (Hach Company, Colorado, USA) to measure free available chlorine levels and compare them to the specified concentration for each product. The Hach DR300 can detect free available chlorine within a range of 0.05 mg/L to 4.00 mg/L. To evaluate the quality of commercial chlorine, two 10-mL samples were prepared using a dilution ratio based on the stock concentration stated on the label by the manufacturer to reach a target concentration of 0.5 mg/L. Commercial chlorine samples were tested on-site at baseline. Two 10-mL undiluted chlorine samples were taken directly from each installed STREAM unit and analyzed at four different time points throughout the study: baseline, month 2, month 4, and endline.
Data on STREAM and commercial chlorine availability and quality were cleaned and entered into an Excel database. Chlorine stock and STREAM production data were used to generate monthly average 0.5% chlorine demand volumes, comparing STREAM actual production values with baseline commercial chlorine volumes across all health facilities and further analyzed by health facility level and region. Similarly, chlorine stockout frequency and duration results were collated, entered into the Excel database, and analyzed by health facility level and region.
An estimated 5-year STREAM total cost of ownership analysis was generated using actual and modeled costs. The analysis included capital costs (e.g., device procurement and spare parts), introductory costs (e.g., installation and initial user training), and operational costs (e.g., vinegar and salt consumables, maintenance, repairs, and spare parts). Cost estimates were based on actual expenditures recorded in project documentation, STREAM usage data, material costs, and commercial chlorine prices paid by health care facilities.
Data from STREAM deployments in eight health care facilities in Ghana’s Eastern Region and ten facilities in Uganda—monitored over a 12-month period—were used to estimate usage rates for consumables such as salt, water, and electricity, as well as operational and capital expenses. Requirements for spare parts and maintenance were determined using historical STREAM deployment data and input from the device manufacturer, Aqua Research. Training and maintenance cost estimates were drawn from published sources and guidance from the GHS on maintenance expenses for comparable equipment.
The costs of commercial chlorine and water were obtained from expense receipts at each health facility during the baseline assessment. These figures were used to determine the cost per liter of 0.5 mg/L commercial chlorine solution.
A per-liter cost of STREAM chlorine was generated using several inputs. These included the capital cost of the STREAM; shipping and customs fees; supplies required for chlorine production (e.g., wooden spoons, measuring cups, 20-liter buckets, and jerry cans); and the recurring costs of consumables, including salt, water, electricity, chlorine test strips, and vinegar. Electricity and water costs were estimated by collecting the unit costs per kilowatt and per liter, respectively, from the national electricity and water utility service providers. These expenses were used to model production and costs over a 5-year period. All cost analyses were conducted in US dollars for the year 2023, using a currency exchange rate of GHS 12.64 per $1.00.
The STREAM total cost of ownership analysis focused on financial costs, excluding opportunity costs associated with existing resource use. The analysis included capital costs, as well as recurrent and operational costs such as those for maintenance, repair, and spare parts. All costs were extrapolated over a 5-year period and compiled from project records in Excel. A flat-rate repair labor cost was included for personnel expenses, but no other personnel costs were considered in the analysis. While personnel costs were excluded from the analysis due to high variability across settings, it is worth noting that the time required for preparing brine and operating the device is expected to be similar to the time needed for diluting commercially procured chlorine, and therefore may not represent a significant additional burden, even in understaffed facilities. The costs of ongoing technical support were estimated using Ghana Ministry of Health/GHS staff rates, assuming similar personnel would continue these activities after project completion.
Materials
The Aqua Research STREAM Disinfectant Generator is an on-site chlorine generator designed to produce a consistent flow of 0.5% mg/L mixed-oxidant solution comprised primarily of sodium hypochlorite with trace amounts of chlorine dioxide, dissolved ozone, hydrogen peroxide, and oxygen (Fig 1). Compared to conventional sodium hypochlorite solutions, mixed-oxidant solutions have demonstrated greater effectiveness in neutralizing viruses, bacteria, and protozoa [19]. The raw materials required to produce the sodium hypochlorite solution are salt and water constituted into a brine solution.
Republished from PATH under a CC BY license, with permission from PATH, original copyright 2023.
Manufactured in New Mexico, USA, the STREAM device operates using basic inputs—salt, water, and electricity—to generate chlorine on-site as needed. It is compatible with multiple power sources, including 110/220 VAC and 12 VDC from a car battery or solar panel. The STREAM holds several regulatory approvals, such as CE marking (No. ES151124043E) for its core reaction cell technology and compliance with the International Electrotechnical Commission’s safety standards for electrical equipment used in measurement, control, and laboratory settings (IEC 61010–1:2010). Independent, certified laboratories have verified the chlorine concentration of the STREAM’s solution, including the Ethiopian Conformity Assessment Enterprise (04/02/2022, ES 877:2022), Uganda National Drug Authority (07/2022, NDA/DLS/FOM/030B), and Essen & Co (03/22/2019). Fig 1 below is an image of the device. Further technical specifications can be found in Table 2.
3. Results
Between June and December 2023, a total of 18 STREAM devices were deployed across 12 health care facilities in Ghana’s Ahafo, Central, and Volta Regions (see Table 2 above for distribution details). These devices were placed in various locations within the facilities, including medical stores (3), laundry units (3), sterilization rooms (2), maternity ward (1), pharmacy (1), dressing unit (1), and reproductive health unit (1). A total of 223 primary users from the 12 health facilities completed the STREAM training.
Reliability
Throughout the study period, STREAM units remained functional for 94.8% of the time. Among the 18 units assessed, 67% (n = 12) operated continuously without any failures, and 33% (n = 6) encountered a single component failure. Notably, no unit experienced more than one component failure. STREAM units operated without any failures for an average of 60.8 days (range 16–96 days). On average, repairs or component replacements for STREAM units were completed within 8 days, which falls within an allowable time frame ranging from 1 to 24 days.
Mechanical failures are defined here as physical components not functioning properly; for example, due to improper assembly by the manufacturer or a defective part from a supplier. Contextual failures are defined here as failures due to environmental factors, such as fluctuations in the power grid or physical water properties that might require extra device cleaning.
Overall, 67% (n = 12) of the STREAM devices were failure-free over the course of this study, and the devices that had failures (due to both mechanical and contextual issues) were able to be repaired and returned to operation.
Mechanical and contextual issues that led to the failures of four STREAM component types are detailed in Table 3.
Chlorine availability and quality
Analysis of chlorine stock cards indicated health care facilities experienced an average of 37 days per year without chlorine. Health centers (n = 7) faced longer stockout periods—nearly 5 times longer (56 versus 11 days) and twice as often per year—as compared to district hospitals (n = 5). There were no reported chlorine stockouts during the study period. (See Table 4).
Significant variations in the availability of commercial chlorine stock were observed among health facilities across regions and health facility levels (Table 5). Health facilities in Central Region received and used significantly more chlorine per month compared to the study health facilities in Ahafo and Volta Regions. Two health facilities in Central Region reported using 11,300–11,697 liters of 0.5 mg/L per month, which was more than 9 times the average compared to the remaining ten study health facilities (see Fig 2). Substantial variations within health facility levels also were seen across all regions.
In total, the STREAM units produced 74,694 liters of 0.5% chlorine disinfectant during the project period, roughly 1,078 liters per month. Half of the study health facilities produced higher monthly 0.5% mg/L chlorine volumes using STREAM devices compared to baseline monthly commercial chlorine volumes. Within these facilities, chlorine volume increased by 763 liters per month. Across the remaining six health facilities, STREAM production reached an average of 876 liters per month, 16% of average chlorine demand (46% when two outliers in Central Region were removed). Reasons for health facilities not being able to produce sufficient STREAM chlorine to address demand included inconsistent production times and cycles, limited awareness among health staff of the STREAM, and extreme chlorine demand levels. The two highest-producing facilities, district hospital 4 and district hospital 5, generated average monthly STREAM chlorine volumes that would nearly address or surpass all but two health facilities in this study.
Concentration testing of commercial chlorine samples at baseline found 83% (n = 10/12) of health facilities were using degraded commercial chlorine (see Fig 3), representing a significant risk to patients and health staff. Health facilities received jerry cans of chlorine labeled as concentrations ranging from 5% to 6% chlorine in volumes of 5–25 liters from the regional medical stores as well as the open market. All samples were taken from chlorine bottles that had not yet reached their expiry date. These commercial chlorine samples were found to contain an average of 45% of the chlorine volume (mg/L) as advertised, resulting in weak and overly diluted disinfectant solutions. In contrast, 100% of the STREAM samples tested in the study’s health care facilities (n = 12) met the STREAM’s target concentration (0.5% ± 0.1% mg/L).
Abbreviation: FCR, free chlorine residual.
Cost and sensitivity analysis
Modeling results indicate that 5-year STREAM chlorine costs would range from $3,700 to $7,900, depending on the volume of chlorine produced. The device capital cost, along with shipping, taxes, and additional items (e.g., spoon, cup, bucket, jerry cans) would comprise roughly 56% to 93% of the total STREAM chlorine cost, and salt and water 7% to 44% of the total STREAM chlorine cost.
Comparing the STREAM chlorine costs using actual production values over the study period with commercial chlorine costs modeled over 5 years, 42% (n = 5) of health facilities had lower per-liter chlorine costs with STREAM devices compared to commercial chlorine costs (see Fig 4). District hospitals would see an average of 17% savings in chlorine supply costs, and health centers would see an average cost increase of 48% (Fig 6). The average cost savings increased to 14% when excluding two high-demand health facilities in Central Region, where unusually low commercial chlorine prices make STREAM chlorine less competitive. At these two facilities, STREAM chlorine is 212% more expensive than commercial chlorine because of the combination of high demand and low commercial chlorine costs. The cost per liter of STREAM chlorine decreases as its production volume increases, reaching breakeven with commercial chlorine at roughly 18,600 liters of 0.5% per year (see Fig 5). Increasing average production volumes to greater than 18,600 liters per device per year would lead to lower chlorine supply costs across eight (67%) of the study health facilities.
Considering annual operational costs of STREAM chlorine production and its consumables (e.g., vinegar, salt, water, and electricity), which do not include training, operational labor, or repair and spare parts costs (as these cannot be generalized across all facilities), compared with commercial chlorine costs (i.e., water and commercial chlorine), the use of STREAM devices would generate a per-liter chlorine cost savings of 64% to 83% per year for health facilities and district hospitals, respectively (see Table 6).
A one-way sensitivity analysis was conducted to evaluate the impact of ±20% changes in input prices (salt, electricity, water, and vinegar) on the unit cost of chlorine produced by the STREAM device (Fig 7). The baseline cost was estimated at $0.0134 per liter at an average annual production volume of 18,000 liters.
The analysis shows that salt is the main cost driver, with chlorine cost ranging from $0.0110 to $0.0160 per liter, or roughly ±19% relative to the baseline. Water had only a minor effect, shifting the cost to between $0.0134 and $0.0135 per liter. Vinegar had a similarly small impact, with costs between $0.0134 and $0.0135 per liter. Electricity showed negligible influence under the tested range, with chlorine cost remaining around $0.0135 per liter.
Overall, the results indicate that salt prices determine the variability in chlorine production costs, while water, vinegar, and electricity contribute little to overall cost changes. These findings suggest that efforts to manage or stabilize chlorine production costs should focus on ensuring reliable and affordable salt supply, while recognizing that the effects of other inputs are limited.
4. Discussion
This study assessed the reliability, operational performance, effect on chlorine availability and quality, and total cost of ownership of a novel electrolytic chlorine generator in public health facilities in Ghana. Overall, the STREAM devices demonstrated a high level of reliability, operating functionally for 94.8% of the study period, with 67% of devices (n = 12) operating continuously without failure. Our study found a higher functionality rate compared to previous STREAM device assessments, in Ugandan health facilities, where 20% of devices operated without any failure [20]. The observed mechanical failures, notably with the circuit boards and controllers, were primarily due to manufacturing processes and similarly observed in Uganda [20]. Issues with the reaction cell and power supplies highlight the effect of contextual and environmental factors (electrical instability [brownouts and power surges], heat stress, and calcium carbonate in source water leading to scaling of electrodes) on long-term functionality of the STREAM device. Similar findings have been recorded in the deployment and evaluation of other electrochlorinator systems in other low- and middle-income countries [21–25]. Overall, the reliability findings suggest that with additional design and manufacturing improvements and site-specific adaptations during installation (e.g., installation of surge protectors, use of rainwater or reverse osmosis systems), the STREAM is well suited for routine use in primary health care settings.
The STREAM device’s effect on chlorine availability in health facilities was mixed. While health facilities operating STREAM devices reported no chlorine stockouts during the study, eliminating the 37 annual days of chlorine stockouts documented in the pre-intervention period, only half of study sites generated equal or greater volumes of 0.5% mg/L chlorine than prior to the study. Our findings on the ability of electrochlorinators to eliminate chlorine stockouts in health facilities is consistent with results from studies in Uganda and Chad [20,26]. The percentage of health facilities (50%) that were able to address or surpass baseline volumes of chlorine available from the STREAM device aligns with results from Uganda, where 50% of district hospitals and health centers met baseline chlorine demand levels [20]. Yet, these results differ from a study involving 103 health centers and rural health posts in Burkina Faso, Ghana, and Liberia, where STREAM devices addressed all chlorine demand [27]. The substantial heterogeneity in chlorine stock levels observed across and among health facilities of the same level may have been a driving factor influencing which facilities were able to address their baseline demand volumes with STREAM devices. Results from our and the aforementioned studies indicate on-site chlorine production may be an effective strategy for addressing frequent interruptions in essential commodity supply chains in lower-level health facilities [28,29]. Thus, while results indicate on-site chlorine production can reduce the dependence on local supply chains and eliminate chlorine stockout periods, greater consideration and analysis are needed to ensure alignment between a health facility’s chlorine demand level and the capacity of the electrolytic chlorine generator.
Our study also revealed a troubling trend in the quality of commercial chlorine used in the study health facilities for disinfection. Analysis of free residual chlorine in commercial chlorine samples found 83% of facilities were using degraded chlorine, which when diluted, produced insufficiently effective disinfection solutions and increased the risk of pathogen transmission in patient areas and cross-resistance of pathogens to antibiotics [30,31]. These findings are similar to findings from previous chlorine quality assessments conducted across several low- and middle-income countries, including in six university hospitals in Benin, where a range of 45% to 100% of commercial chlorine samples tested were found to be degraded [32,33]. In contrast, 100% of the STREAM-produced chlorine samples met the device’s target chlorine concentration, providing a valuable and trusted alternative for health staff. These observations illustrate the potential advantages of in situ chlorine production, limiting the exposure of factors (e.g., ultraviolet light exposure, storage in ambient temperatures exceeding 25ºC, and use of transparent storage containers) than have been shown to cause degradation in chlorine concentration [32,34–36].
The total cost of ownership and breakeven analysis offer valuable insights for the GHS into the financial viability and appropriate health facility level for STREAM devices. The total cost of ownership over 5 years ranged from $3,700 to $7,900 depending on production volume. Per-liter chlorine cost savings were more frequently seen in district hospitals compared to health centers, due to higher chlorine demand in the hospitals. Our results align with prior research that demonstrated operational costs of chlorine generation can generate cost savings compared to commercially available chlorine [37]. Similarly, the potential for cost savings, particularly in district hospitals, aligns with cost analysis results from Uganda, where STREAM devices were estimated to yield a 36.9% reduction in chlorine costs over 5 years [20]. Incentivizing higher utilization of STREAM devices beyond 18,600 liters per year could improve financial viability and expand the geographies for STREAM introduction into health centers and district hospitals in Ghana’s health system. Finally, alternative production and distribution approaches—such as decentralized hub-and-spoke models that have been successfully used for the generation and distribution of alcohol-based hand rub in Uganda—could be explored as mechanisms that align with Ghana’s network of practice approach and utilized to increase chlorine availability in smaller, neighboring dispensaries and community-based health planning and service compounds [38,39]. These findings are directly relevant as the GHS launches scale-up efforts to expand use of the STREAM device in its public health system.
Strengths and limitations
Our study had several strengths and certain limitations. To our knowledge, this is the first study to rigorously examine the reliability and total cost of ownership of an electrolytic chlorine generator in primary health care facilities in Ghana. Findings provide valuable insights for the GHS on installation, monitoring, and maintenance and repair factors that should be considered as additional STREAM devices are installed in health facilities throughout the country. Similarly, our published data on the quality of commercial chlorine have raised concerns among GHS staff regarding the quality of existing chlorine supplies used across the health system and questions about the need for more quality assurance testing.
Several key limitations should be acknowledged. The relatively small number of health facilities participating in the study and the limited geographic coverage of the intervention (3 of Ghana’s 16 regions) may not reflect the unique contextual realities across Ghana and limits the generalizability of our findings. As our findings indicate, there are significant differences across health facilities of the same “level” and the extent of this variation throughout Ghana’s health system remains unknown. While research shows health facilities in Ghana face significant challenges with maintaining continued stock of essential medical commodities due to many of the same factors limiting chlorine availability in our study health facilities (e.g., limited hospital stock management practices, poor road infrastructure, and limited transportation services) [40,41], additional research is needed to better understand chlorine supply chain gaps and the potential value of the STREAM device in those locations. The total cost of ownership analysis did not capture certain labor costs (i.e., training, supervision, operation, and repair). Expanding this analysis to include the full spectrum of costs related to installation, operation, chlorine use, maintenance, and spare parts and repair services would provide a more comprehensive total cost of ownership of the STREAM device. Additionally, further quantifying STREAM and commercial chlorine production requirements (e.g., device operation, stock management, time, and dilution factors), combined with qualitative psychosocial factors related to the elimination of STREAM dilution for disinfection and implications of on-site chlorine production for IPC factors should be considered. Finally, health facilities may have relied on commercial chlorine products to supplement STREAM chlorine (without tracking it on stock cards) to address peak demand periods and/or relied solely on commercial chlorine products while STREAM units were nonfunctional. Additional investigation and improved facility-level monitoring is needed for future implementation.
Future implementation research approaches could be beneficial in evaluating how well the device integrates into the health system. This could include assessing its role and pairing STREAM introduction with IPC training, piloting novel STREAM chlorine production and distribution models that address chlorine availability gaps within the health sector through a financially viable approach, and exploring the use of STREAM chlorine for water treatment as well as disinfection.
Aligned with findings from a recent systematic review on interventions aimed at improving water supply, sanitation, and hand hygiene in health care settings in low- and middle-income countries, further studies should explore the impact of consistent availability of cleaning agents—particularly chlorine—on reducing HAIs. Research on chlorine’s effectiveness both as a stand-alone intervention and as part of a broader IPC strategy in these settings would be particularly valuable.
5. Conclusions
This study demonstrates the potential benefits and limitations of the STREAM device in addressing persistent challenges related to chlorine availability and quality in public health facilities in Ghana. Findings on device reliability and cost implications provide actionable evidence that can inform the GHS’s national STREAM introduction strategy and advance IPC practices and objectives described in Ghana’s National Infection Prevention and Control Strategy (2024–2028). To maximize impact, future efforts should prioritize optimized device deployment strategies and expanded operational models—such as a decentralized hub-and-spoke production and distribution model—that enable sustained use, strengthen troubleshooting and maintenance protocols, and support facilities in scaling chlorine production to achieve cost efficiency. Addressing contextual barriers, such as power supply variability and water quality considerations, will further strengthen the sustainability and impact of STREAM devices in health care settings. Ultimately, integrating the STREAM into routine health service delivery has the potential to improve the quality of IPC practices and contribute to safer environments for patients and health care workers.
Supporting information
S1 Table. Summary of commercial chlorine stockcard.
https://doi.org/10.1371/journal.pwat.0000383.s001
(XLSX)
S2 Table. Commercial and STREAM chlorine quality.
https://doi.org/10.1371/journal.pwat.0000383.s002
(XLSX)
S3 Table. STREAM monthly chlorine production.
https://doi.org/10.1371/journal.pwat.0000383.s003
(XLSX)
S4 Table. STREAM and commercial breakeven analysis.
https://doi.org/10.1371/journal.pwat.0000383.s004
(XLSX)
S5 Table. STREAM 5-year total cost of ownership.
https://doi.org/10.1371/journal.pwat.0000383.s005
(XLSX)
Acknowledgments
This research was conducted with support from the Ghana Health Service’s Institutional Care Division, notably Dr. Ofori-Boadu, Mrs. Gloria Ntow-Kummi, and Dr. Mary Ashinyo.
References
- 1. Allegranzi B, Bagheri Nejad S, Combescure C, Graafmans W, Attar H, Donaldson L, et al. Burden of endemic health-care-associated infection in developing countries: systematic review and meta-analysis. Lancet. 2011;377(9761):228–41. pmid:21146207
- 2.
World Health Organization (WHO). Report on the burden of endemic health care-associated infection worldwide. Geneva: WHO; 2011. Available from: https://apps.who.int/iris/handle/10665/80135
- 3.
World Health Organization (WHO). Health-care associated infections [fact sheet]. Geneva: WHO; 2016.
- 4.
Hutton G, Chase C, Kennedy-Walker RJ. Costs of healthcare-acquired infections due to inadequate water, sanitation and hygiene (WASH) in healthcare facilities in Ghana. Policy Research Working Paper No. WPS 10708. Washington, DC, USA: World Bank Group; 2024. Available from: https://washmatters.wateraid.org/sites/g/files/jkxoof256/files/2024-04/Costs-healthcare-acquired-infections-Ghana.pdf
- 5.
World Health Organization (WHO). Global report on infection prevention and control. Geneva: WHO; 2022. Available from: https://apps.who.int/iris/handle/10665/354489
- 6. Schreiber PW, Sax H, Wolfensberger A, Clack L, Kuster SP, Swissnoso. The preventable proportion of healthcare-associated infections 2005-2016: Systematic review and meta-analysis. Infect Control Hosp Epidemiol. 2018;39(11):1277–95. pmid:30234463
- 7. Storr J, Twyman A, Zingg W, Damani N, Kilpatrick C, Reilly J, et al. Core components for effective infection prevention and control programmes: new WHO evidence-based recommendations. Antimicrob Resist Infect Control. 2017;6:6. pmid:28078082
- 8. Umscheid CA, Mitchell MD, Doshi JA, Agarwal R, Williams K, Brennan PJ. Estimating the proportion of healthcare-associated infections that are reasonably preventable and the related mortality and costs. Infect Control Hosp Epidemiol. 2011;32(2):101–14. pmid:21460463
- 9.
Organization for Economic Co-operation and Development. Stemming the superbug tide: just a few dollars more. Paris: OECD Publishing; 2018.
- 10.
Ghana Ministry of Health. MoH launches National Infection Prevention and Control Strategy. n.d. [Cited 2025 March 17] Available from: https://www.moh.gov.gh/moh-launches-national-infection-prevention-and-control-strategy/
- 11. Oppong TB, Amponsem-Boateng C, Kyere EKD, Wang Y, Gheisari Z, Oppong EE, et al. Infection Prevention and Control Preparedness Level and Associated Determinants in 56 Acute Healthcare Facilities in Ghana. Infect Drug Resist. 2020;13:4263–71. pmid:33262620
- 12. Opollo MS, Otim TC, Kizito W, Thekkur P, Kumar AMV, Kitutu FE, et al. Infection Prevention and Control at Lira University Hospital, Uganda: More Needs to Be Done. Trop Med Infect Dis. 2021;6(2):69. pmid:34062871
- 13.
US Agency for International Development (USAID). Improving the quality of health care services by strengthening IPC at centers of excellence: Technical report. Washington, DC, USA: USAID; 2021.
- 14. Cronk R, Bartram J. Environmental conditions in health care facilities in low- and middle-income countries: Coverage and inequalities. Int J Hyg Environ Health. 2018;221(3):409–22. pmid:29352706
- 15.
PATH and Ghana Health Service Observational Study – Unpublished project report, 2021.
- 16. Boata M, Johnson R., Yaa O, Boakye D, Amankwah F, Owusu F, et. al. Efficacy of disinfectants commonly used in some medical facilities in Kumasi, Ghana. J Sci Technol. 2023;41: 65-–81. Available from: https://www.ajol.info/index.php/just/article/view/246176
- 17. Stephen TO, Tahiru M, Anthony MS. Long-term effectiveness of commonly used hospital disinfectants against Pseudomonas aeruginosa and Staphylococcus aureus. Afr J Microbiol Res. 2021;15(7):396–400.
- 18. National Institute of Standards and Technology (NIST) and Semiconductor Manufacturing Technology consortium. e-Handbook of statistical methods. Gaithersburg, Maryland, USA: NIST; 2003. Available from:
- 19. Venczel LV, Arrowood M, Hurd M, Sobsey MD. Inactivation of Cryptosporidium parvum oocysts and Clostridium perfringens spores by a mixed-oxidant disinfectant and by free chlorine. Appl Environ Microbiol. 1997;63(4):1598–601. pmid:9097455
- 20. Drolet A, Mugumya T, Hsu S, Izudi J, Ruhweza M, Mugisha E, et al. Performance and acceptability of the STREAM Disinfectant Generator for infection prevention and control practices in primary health care facilities in Uganda. Antimicrob Resist Infect Control. 2024;13(1):77. pmid:39014488
- 21. Howitt P, Darzi A, Yang G-Z, Ashrafian H, Atun R, Barlow J, et al. Technologies for global health. Lancet. 2012;380(9840):507–35. pmid:22857974
- 22. Malkin RA. Design of health care technologies for the developing world. Annu Rev Biomed Eng. 2007;9:567–87. pmid:17430083
- 23. Bhattacharya M, Bandyopadhyay K, Gupta A. Design of a cost-effective electrochlorination system for point-of-use water treatment. Environmental Engineering Research. 2020;26(5):200437–0.
- 24. García-López I, Arenas LF, Hereijgers J, Águeda VI, Garrido-Escudero A. Drinking Water Electrochlorination in a Single-Pass Biomimetic Flow Cell with Zero Salt Dosing. ACS Sustainable Chem Eng. 2024;12(8):3130–41.
- 25. Ferro S. Challenges in Designing Electrochemical Disinfection Systems for Reducing Microbial Contamination in Drinking Water Distribution Networks. Water. 2025;17(5):754.
- 26. Campos Nougeira R, Nigro M, Veuthey J, Tigalbaye C, Bazirutwabo B, Fam Thior MD, et al. Can locally produced chlorine improve water sanitation & hygiene (WASH) indicators in health care facilities (HCF) in rural Chad?. Health Sci Dis. 2022;22:11.
- 27.
Fofie F. No breach in chlorine supply for health care facilities in Burkina, Ghana, and Liberia. May 2021.
- 28. Frimpong G, Ofori-Kwakye K. Access to Essential Medicines in Ghana: A Survey of Availability of Children’s Medicines in Medicine Outlets in the Ashanti Region. J App Pharm Sci. 2016;:020–8.
- 29. Leung N-HZ, Chen A, Yadav P, Gallien J. The Impact of Inventory Management on Stock-Outs of Essential Drugs in Sub-Saharan Africa: Secondary Analysis of a Field Experiment in Zambia. PLoS One. 2016;11(5):e0156026. pmid:27227412
- 30. Nam J-H, Yoo JS. Sublethal Sodium Hypochlorite Exposure: Impact on Resistance-Nodulation-Cell Division Efflux Pump Overexpression and Cross-Resistance to Imipenem. Antibiotics (Basel). 2024;13(9):828. pmid:39335002
- 31. Aljuwayd M, Malli IA, Ricke SC, Kwon YM. A Sublethal Concentration of Chlorine Induces Antibiotic Resistance in Salmonella via Production of Reactive Oxygen Species. Applied Microbiology. 2024;4(2):745–52.
- 32. Soclo Dansi SE, Degbey CC, Kpozehouen A, Gaffan N, Agbokou AD, Tapha O, et al. Stability and efficacy of chlorinated disinfectants in beninese hospitals: issues for the prevention and control of infections and antibiotic resistance. Preprints. 2025. Available from:
- 33. Lantagne DS. Viability of commercially available bleach for water treatment in developing countries. Am J Public Health. 2009;99(11):1975–8. pmid:19762657
- 34. Gow CK, Weinhouse C, Johnson GO, Saunders KE. Stability of Free Available Chlorine Levels in Dilute Sodium Hypochlorite Solutions over a 6-Week Period. J Am Assoc Lab Anim Sci. 2022;61(2):181–7. pmid:35109965
- 35. Chaúque BJM, Rott MB. Photolysis of sodium chloride and sodium hypochlorite by ultraviolet light inactivates the trophozoites and cysts of Acanthamoeba castellanii in the water matrix. Journal of Water and Health. 2020;19(1):190–202.
- 36. Ahmad S, Almehmadi M, Janjuhah HT, Kontakiotis G, Abdulaziz O, Saeed K, et al. The Effect of Mineral Ions Present in Tap Water on Photodegradation of Organic Pollutants: Future Perspectives. Water. 2023;15(1):175.
- 37. Naranjo-Soledad A, Smesrud L, Bandaru SRS, Hernandez D, Mehare M, Mahmoud S, et al. Low-cost, local production of a safe and effective disinfectant for resource-constrained communities. PLOS Glob Public Health. 2024;4(6):e0002213. pmid:38917188
- 38. Tusabe F, Nanyondo J, Lozier MJ, Kesande M, Tumuhairwe O, Watsisi M, et al. Improving Access to WHO Formulations of Alcohol-Based Hand Rub in Healthcare Facilities: A District-Wide Approach. Am J Trop Med Hyg. 2023;109(1):191–200. pmid:37188343
- 39.
Ghana Ministry of Health. National healthcare quality strategy: Revised edition (2024–2030). Accra, Ghana: Ministry of Health. 2024. https://www.moh.gov.gh/wp-content/uploads/2025/05/National-Health-Quality-Strategy-2024-2030.pdf
- 40. Atiga O, Walters J, Pisa N. Challenges of medical commodity availability in public and private health care facilities in the Upper East Region of Ghana: a patient-centered perspective. BMC Health Serv Res. 2023;23(1):719. pmid:37393292
- 41. Boakye G, Gyedu A, Stewart M, Donkor P, Mock C, Stewart B. Assessment of local supply chains and stock management practices for trauma care resources in Ghana: a comparative small sample cross-sectional study. BMC Health Serv Res. 2021;21(1):66. pmid:33441147