Pilot implementation of "Outsourced Oxygen to the Bedside" models in five countries: a mixed methods impact assessment

A mixed-methods assessment of "Outsourced Oxygen to the Bedside" pilots across five countries reveals that private provider models significantly improved the functionality of oxygen equipment and healthcare worker knowledge, though successful implementation requires better local tailoring and capacity building to address remaining service gaps.

The Gist

Imagine a hospital ward as a busy kitchen during a dinner rush. The chefs (doctors and nurses) are trying to save lives, but they can't cook the most important dish—oxygen therapy—because the stove is broken, the gas tank is empty, or they don't know how to turn the knob.

This paper is about a new experiment to fix that kitchen. The researchers tested a model called "Outsourced Oxygen to the Bedside" (O2B). Think of this like hiring a specialized private delivery service to not just drop off a gas tank, but to manage the entire gas system for the hospital.

Here is the story of what they found, broken down simply:

1. The Problem: The "Broken Stove" Crisis

Before this experiment, many hospitals in Kenya, Nigeria, India, Tanzania, and Uganda were struggling.

• The Equipment Graveyard: They had oxygen machines, but they were often broken, like a car with a flat tire sitting in a junkyard.
• The Empty Tank: When they did have tanks, they were often only half-full because the delivery trucks were slow or unreliable.
• The Missing Tools: Sometimes they had oxygen but no masks or tubes to connect it to the patient.
• The Fear: The staff were scared to use the equipment because they didn't know how to fix it if it broke, so they often just sent sick patients away to bigger hospitals.

2. The Solution: The "All-In-One" Service

The researchers tested five different "delivery services" (private companies) in these countries. Instead of the hospital buying a machine and hoping it works, they paid a company to provide a bundle of services:

• The Hardware: New oxygen concentrators (machines that pull oxygen from the air) and cylinders.
• The Maintenance: A "swap-and-go" policy. If a machine breaks, the company swaps it for a working one immediately and fixes the broken one later. No waiting around.
• The Training: Teaching the nurses how to use the new tools.
• The Monitoring: Keeping track of how much oxygen is used so they never run out.

3. The Results: A Much Happier Kitchen

The researchers went into 28 hospitals and checked the "kitchens." Here is what they found:

• The Machines Worked: This was the biggest win.

  • Old Machines: Only about 30% of the old oxygen cylinders and 20% of the old machines were actually working.
  • New Service Machines: 64% of the new cylinders were full and ready, and a whopping 95% of the new machines were working perfectly.
  • Analogy: It's like switching from a kitchen where half the stoves are broken to one where almost every stove is hot and ready to cook.

• The Staff Felt Safer:

  • Before, nurses were hesitant. "If I turn this on and it breaks, who do I call?"
  • After, they felt confident. "I know if this breaks, the delivery guy will swap it in 10 minutes." This confidence meant they treated more sick children right there in the hospital instead of sending them away.

• The "What's Included?" Confusion:

  • There was a bit of a mix-up. The companies thought they were providing a full "all-you-can-eat" buffet (training, repairs, masks, etc.), but some hospital staff thought they were only getting the gas tank.
  • Analogy: It's like ordering a "Premium Pizza Package" that includes a drink and dessert, but the customer only sees the pizza and thinks, "Where's my soda?" The staff didn't always realize all the extra help they were getting.

4. The Catch: It's Not Perfect Yet

While the "delivery service" fixed the broken stoves, there are still some gaps:

• Training Gaps: Even with the new machines, some staff still needed more practice. The training provided was good, but because staff change jobs often (high turnover), they need constant "refresher courses."
• Not Every Ward Got It: In some places, the service only covered the baby ward, leaving the adult ward with the old, broken equipment.
• The "Whole Facility" Dream: Some hospitals wanted the service to cover everything in the building, but the pilots were sometimes limited to just specific areas.

The Bottom Line

Think of the O2B model as hiring a professional property manager for your house instead of trying to fix the plumbing yourself.

• Before: You bought a pipe, it leaked, you didn't know how to fix it, and your house flooded.
• After: You pay a service. They bring the pipe, they fix it instantly if it leaks, and they teach you how to turn the water on.

The verdict: This approach works! It makes oxygen reliable and available at the patient's bedside. However, to make it a permanent solution for the whole world, the "service packages" need to be clearer, the training needs to be continuous, and the companies need to be ready to cover the whole hospital, not just one room.

In short: Giving hospitals a reliable "oxygen subscription" with a dedicated repair team saves lives by ensuring the life-saving gas is actually there when the patient needs it.

Technical Summary

1. Problem Statement

Medical oxygen is an essential medicine, yet an estimated 60% of the global population lacks access to reliable, affordable, and quality oxygen services. This gap is most severe in Sub-Saharan Africa and South Asia. While the COVID-19 pandemic spurred investments in oxygen equipment, critical weaknesses remain in distribution systems, supply chains for spare parts, biomedical engineering capacity, and healthcare worker (HCW) training. Consequently, even when equipment is present, it is often non-functional or interrupted.

The paper investigates the "Outsourced Oxygen to the Bedside" (O2B) model, a subset of "Oxygen as a Service" (O2aaS). In this model, a private provider delivers a bundled package of services (equipment, maintenance, repair, and sometimes training/consumables) to health facilities for a single price, rather than the facility managing oxygen as a commodity. The study aims to determine if these pilots successfully improve reliable oxygen access at the patient bedside and to identify the mechanisms driving these outcomes.

2. Methodology

The study employed a sequential explanatory mixed-methods design to evaluate five O2B pilot projects across Kenya, Nigeria, India, Tanzania, and Uganda (conducted September 2024 – January 2025).

• Study Sites: 28 facilities were selected from a total of 80 participating in the pilots. These included primary, secondary, and tertiary facilities (public, private, and faith-based) across diverse health system contexts.
• Quantitative Component:
  • Facility Audit: A cross-sectional audit assessed facility contexts, equipment functionality (cylinders, concentrators, pulse oximeters), and usability.
  • HCW Survey: 179 healthcare workers completed a knowledge survey on oxygen therapy and pulse oximetry.
  • Comparison: Functionality and usability were compared between equipment provided by O2B providers versus non-O2B equipment already present in the wards.
• Qualitative Component:
  • Interviews: 59 semi-structured interviews were conducted with facility managers, HCWs, and policy makers across 20 facilities.
  • Focus: Explored experiences with the O2B pilots, understanding of service packages, maintenance responsiveness, and training effectiveness.
• Analysis: Quantitative data used chi-square and t-tests for differences in functionality and knowledge scores. Qualitative data was analyzed using a pragmatic framework approach (Gale et al.) and triangulated with quantitative findings using implementation science frameworks (Proctor's outcomes and CFIR).

3. Key Contributions

• Empirical Evidence on O2aaS: This is one of the first rigorous mixed-methods evaluations of O2B models in low-resource settings, moving beyond theoretical proposals to assess real-world implementation.
• Mechanism Identification: The study identifies specific mechanisms (equipment functionality, maintenance reliability, and staff confidence) through which O2B models improve access, rather than just measuring availability.
• Contextual Variability: It highlights the heterogeneity of O2B packages (e.g., whole-facility vs. ward-specific, solar-powered vs. grid-dependent) and the importance of tailoring services to local pre-existing gaps.
• Gap Analysis: It reveals a critical disconnect between what providers intend to offer and what facility staff perceive they are receiving, particularly regarding training and maintenance services.

4. Key Results

A. Equipment Functionality and Usability
The O2B-provided equipment was significantly more functional and usable than non-O2B equipment:

• Cylinders: 64.5% of O2B cylinders were usable vs. 40.9% of non-O2B ($p < 0.001$).
• Concentrators: 95.0% of O2B concentrators were functional vs. 25.7% of non-O2B ($p < 0.001$). The primary reason for non-functionality in non-O2B units was the inability to turn them on.
• Pulse Oximeters: 84.0% of O2B devices were functional vs. 70.0% of non-O2B ($p = 0.172$, not statistically significant but higher).
• Pediatric/Neonatal Specifics: Only facilities with O2B equipment had age-appropriate pulse oximeter probes and flow splitters available.

B. Healthcare Worker (HCW) Knowledge and Training

• Training Coverage: Only 21.8% (39/179) of HCWs reported receiving training from O2B providers.
• Knowledge Scores: HCWs who received O2B training had slightly higher mean knowledge scores (15.3/24) compared to those who did not (13.9/24) ($p = 0.002$). The difference was primarily driven by pulse oximetry knowledge.
• Confidence: Qualitative data indicated that reliable oxygen supply increased HCW confidence to treat critically ill patients on-site, reducing unnecessary referrals. However, knowledge gaps remained, and staff turnover threatened the sustainability of training gains.

C. Service Delivery and Perception

• Maintenance: A key success factor was proactive maintenance and "swap-and-go" policies (replacing broken equipment immediately). However, interviewees reported inconsistent experiences, with some facing delays.
• Misalignment: There was significant confusion among staff regarding the scope of the O2B package. Some staff were unaware that training or specific consumables were included, while others mistakenly believed the service covered areas not intended by the provider.
• Supply Chain: O2B providers successfully mitigated pre-existing supply chain issues (e.g., half-filled cylinders, unreliable refilling) by guaranteeing full cylinders and reliable delivery.

5. Significance and Implications

• Proof of Concept: The O2B model demonstrates potential to solve the "last mile" problem of oxygen delivery by bundling equipment with essential services (maintenance, training, supply chain management).
• Critical Success Factors: The study suggests that the reliability of maintenance and repair services is as critical as the equipment itself. Without functional equipment, access is illusory.
• Implementation Challenges:
  • Communication: There is a need for clearer communication between providers and facilities to align expectations regarding service scope (e.g., training, consumables).
  • Capacity Building: While equipment availability improved, HCW knowledge gaps persist. O2B providers may need to evolve into "oxygen mentors" to provide continuous on-the-job training.
  • Scalability: Models must be adaptable; a "whole facility" approach may be needed in some contexts, while targeted ward-specific solutions work in others.
• Future Directions: The authors call for further economic evaluations to determine cost-effectiveness and long-term sustainability without donor support. They also emphasize the need for systems to forecast oxygen demand based on improved HCW confidence and utilization rates.

Conclusion:
The O2B pilots successfully improved reliable medical oxygen access at the bedside primarily through superior equipment functionality and maintenance services. However, for these models to be sustainable and equitable, they must address gaps in HCW capacity, clarify service packages, and adapt to local health system contexts.