To improve TB infection prevention and control (IPC) in health facilities, CO₂ monitors are installed in to monitor carbon dioxide levels as a proxy for TB transmission risk. Hospital waiting rooms, outpatient clinics, emergency departments and inpatient wards are often ventilated at levels well below those recommended for the control of TB transmission.

CO₂ monitors are utilised to evaluate environmental controls and ensure efficacy of infection control measures in health facilities real-time. Warning alarms prompt the healthcare workers to improve ventilation in waiting areas through opening windows or moving people to an outdoor area.

Through a partnership with the USAID Tuberculosis South Africa Project, the Council for Scientific and Industrial Research (CSIR) has developed GPRS-enabled CO₂ monitors that can serve as a real-time alarm for the management and operational response of indoor areas with high potential for airborne disease transmission. The differential between indoor and outdoor CO₂ levels can be measured and used to estimate the ratio of rebreathed air in a space and thereby the risk of airborne disease transmission. The monitor can also be used as a simple data logger to record indoor air quality metrics such as CO₂, temperature and humidity. 

Rationale and Relevance to TB Infection Prevention and Control

CO₂ is produced by people when they breathe. Each exhaled breath by an average adult contains 35,000 to 50,000 parts per million (ppm) of CO₂ – 100 times that of the typical concentration in the outside air. By measuring the indoor CO₂ concentration in different rooms of a health facility as a proxy, the ratio of rebreathed air can be used to estimate the risk of airborne disease transmission in a given space. CO₂ Threshold values can be established such that less than 1 transmission occurs for each exposure event. By setting such CO₂ threshold values, health care workers can be alerted when CO₂ concentrations indicate epidemic levels of indoor transmission. Action can then be taken to either increase ventilation or to remove people from the space.

Benefits and Advantages

The use of CO₂ monitors confers numerous benefits and advantages as part of a complete, holistic IPC strategy. These benefits include:

• Measuring the ventilation in a given space (using CO₂ concentration as a proxy) permits the quantitate evaluation other environmental control interventions
• Improved information for facility management and additional data for real time decision making at the ward and facility level
• Targeted, data-driven IPC responses for high risk spaces and periods identified in high-burden facilities
• Relatively simple and low cost setup and installation of CO₂ monitors and clear path for scale up with automatic data collection and storage in a secure database

Testimonials from Health Care Workers

Guidelines for Monitor Placement

The following guidelines are used for placing monitors in the facilities:

• The device should not be directly breathed on during deployment or operation as breath can give very high false readings. Thus, the device should be installed above or below the direct normal breathing zone in the space by about 300mm.
• Do not place the device on working desks or counters as this may attract attention where the device could be breathed on.
• The device should ideally be placed near ventilation outlets but away from ventilation inlets.
• If stagnancy zones are expected in the room and people are expected to congregate in these areas, the device should be placed near these areas.
• The device can be stuck to the wall with double sided tape, but the tape will then need to be removed with a solvent after retrieval without damaging the plaster or paint. Ideally the device can be fixed using 5x35mm wall plugs
• If the device needs to be deployed in rooms where occupants might be disturbed by the indicator light, or where the light may attract unwanted attention it can be turned off. To switch-on the indicator light, press the “LED” button, the device can be paused using the “PAUSE” button and logging can be resumed using the “CO₂” buttons in the terminal app. 

Data Collection

CO₂ monitors collect CO₂ concentration (in parts per million [ppm]), temperature, relative humidity and battery voltage data in real time (the average of 60 readings every 15 minutes). These readings are collected and stored locally on an SD card, and are subsequently sent via cellular network to a secure, centralized server for storage and analysis. CO₂ monitors require an external power source for operation and thus collect voltage to understand whether a power outage occurred during the day.

CO₂ Alerts and Alarms

The CO₂ monitors can be configured to have two thresholds: an alert limit (Typically 550 ppm or more) and an alarm at 1100 specification (OOS) reading to health care workers via automatic SMS/text message, prompting them to decant those areas that are congested or poorly ventilated. When the average CO₂ level exceeds the alert limit for at least one hour, the device triggers an SMS alarm to health care workers. If the average CO₂ level exceeds the alarm level for any 15-minute period, the device triggers an SMS alarm to health care workers.

SMS alerts and alarms contain the location of the device, time, average CO₂ level, minimum CO₂ level, maximum CO₂ level, temperature, Risk Index (%) and battery charge level. Each device sends SMS alerts individually to one IPC/FAST champion in the facility. IPC/FAST champions are coached to recognize that when the minimum CO₂ level is exceeded, they must alert the health care workers to either improve ventilation or decant the affected areas. 

On-Demand Data Dashboards

The on-demand dashboard through the IPConnect suite of applications allow facility managers, program managers and other key stakeholders to look both at real time CO₂ data and at historical trends to facilitate immediate action planning for improved IPC over time.

Data Analysis Procedures

Initial data analysis procedures include examining the number of OOS events in a given day, the % of time in a day during which the CO₂ monitors recorded an OOS event, and the R0e% in a given day (that is, the probability of TB transmission given the CO₂ concentration). 

Average OOS Events

This value is calculated by examining the number of times a given CO₂ monitor registered an OOS (typically above 550 ppm) value and taking the sum for the day. Currently this value is averaged over the quarter at the individual CO₂ monitor level. Future work in IPConnect will allow the user to select the duration of time over which to average the value (i.e. OOS events in a day, week or month).

Daily OOS Hours (%)

This value is calculated by examining, for each time a CO₂ monitor registers an OOS (above 550 ppm) value, the time the monitor continues to register OOS values. The total time a given monitor records OOS values for the given day is summed and taken as a percentage of the entire day. This value is currently averaged over the quarter at the individual CO₂ monitor level. Future work in IPConnect will allow the user to select the duration of time over which to average the value (i.e. average OOS hours in the given day, week or month).

R0e

This value, the environmental reproduction number, is described in detail by van Reenen in “CO₂ Alarm Limit for Airborne Disease Transmission” (2017). In summary, this value represents the number of secondary cases (in this case, of TB) derived from a primary case (of TB). If this value is less than one, the disease-free equilibrium is maintained, and transmission would not occur. R0e values above one, then, suggest onward TB transmission. To calculate the R0e the number of room occupants is required. For this analysis, the number of room occupants is assumed to be 100 and the number of infectors to be 1 in 100. This allows the R0e to represent the percentage of transmission events (R0e%) per number of exposed room occupants. It is important to caution here that transmission events and progression to TB disease should not be conflated. The R0e% reported therefore remains useful as index of risk alone until a susceptibility factor can be introduced. For this analysis, an environment with R0e% of 1.0% or greater is considered to contribute towards outbreak conditions.