Factors to consider in the safe design of intensive care units – Part 1: historical aspects and ventilation systems

Abstract

Background

Evidence linking the role of ventilation systems in transmission of infection to patients in intensive care units has increased in recent years.

Aims

This research-based commentary set out to identify the historical aspect of intensive care unit design, current problems and some potential solutions with respect to ventilation systems.

Methods

Databases and open source information was used to obtain data on the historical aspects and current guidance in ICU, and the authors experiences have been used to suggest potential solutions to ventilation problems in ICU.

Findings

The authors found a number of problems with ventilation in ICU to which there has not been a cohesive response in terms of guidance to support users and designers. The resultant void permits new projects to proceed with suboptimal and designs which place patients and staff at risk.

Discussion

The NHS is now at the start of major new investments in healthcare facilities in England and this together with the end of the antibiotic era mandates new guidance to address these major concerns.

Introduction

Historical and evolution of ICU design to control microbial transmission

The origins of the Intensive Care Unit (ICU) come from the concept of progressive patient care (Thoms, 1962). Clinicians in the 1950s described five basic elements of care, the first of which was the specialised intensive treatment for the critically ill (Thoms, 1962). Care for these patients had traditionally taken place in a small number of beds within an existing ward (Robinson, 1966). Early ICU’s consisted of six beds surrounded by a centralised nursing area where equipment and drugs were stored with some units having single bed cubicles (Hamilton, WK, n.d.). Each hospital had to work out its own plan in relation to its own needs and the patient population served was emphasised (Thoms, 1962). Workshops developed the concept further and focused on cross infection, with every patient described as ‘an exercise in bacteriologic control’ (Hamilton, WK, n.d.). Routes of transmission including airborne were discussed, and removal of pathogens by an ultraviolet source or by ventilation systems was proposed. Air changes were cited as a requirement not just for comfort but for dilution of pathogens with 10 ACH/hr described as optimal (Hamilton, WK, n.d.).

Some of the earliest UK ICUs were constructed in the 1960s (Robinson, 1966). Early infection control measures included single rooms, designated nurses for each patient and removal of jackets and white coats on entry with gowns and footwear for each room. Rooms contained the required instruments, drugs and solutions and movement between rooms was minimised and ethylene oxide sterilisation of ventilators was undertaken with a positive pressure air gradient from the rooms to the corridor to reduce infection risk (Robinson, 1966). Two options were proposed 1) a normal ward type accommodation was provided over a large area with a limited number of isolation rooms and in the second, the design was based almost on single room accommodation with a stepdown area for patients whose nursing requirements were less (Robinson, 1966).

Over time, seven types of ICU layout were described, including the racetrack variety (in 12/19 ICUs) where the services were in the centre and patient beds in the perimeter (James and Tatton-Brown, 1986; Rashid, 2006). Advantages to the racetrack layout included maximisation of perimeter walls, natural light in patient rooms and reduced walking distances for staff. In one study, 32% (six of 19) had no isolation rooms (Rashid, 2006); and in another, 85% of ICUs (33 units) did not have a single room for isolation with 60 % having fewer than one washbasin per bed space, and the authors called for alterations to enable adequate hand washing to be carried out and for source isolation of infected patients (Inglis et al., 1992).

A task force published recommendations on minimal requirements for ICUs including an essential ratio of isolation rooms to common rooms of 1–2/10, two sinks in each room, elbow or foot operated taps and self-sterilising heat traps with storage areas 30m (max) away from the patient area and 5 m² per bed for consumables and equipment (Ferdinande, 1997). Recommendations were updated in 2011, with single rooms to reduce cross contamination and minimise stress from noise and activities, with some to be equipped as isolation rooms (Valentin et al., 2011). A separate circuit for evacuation of contaminated material in the room was also recommended as well as more detailed recommendations for waste disposal including a fluid disposal or bedpan flushing device available in each patient room or as part of an adjacent toilet (Thompson et al., 2012). However, bedpan washers may generate infectious aerosols so barriers or sealed models should be used to protect staff from exposure. Where fluid disposal was not available in the room, it should be provided in close proximity in the corridor (Thompson et al., 2012).

Methodology

A narrative review was performed using relevant search terms: intensive care unit OR critical care unit AND design OR design guidelines OR design criteria OR ventilation OR ventilation specification OR ventilation systems. The databases employed were PubMed, CINAHL, CDSR, DARE and EMBASE from ‘Jan 2000 to April 2022’ for guidelines, reviews and original articles detailing recommendations for intensive care settings. Published data were checked for additional references and duplicate publication.

Ventilation aspects of ICU design

Ventilation specification (HBN 04–02) guidance is available for ICUs (NHS England, 2021a); however, ventilation specification details are contained within an appendix (S) HTM 0301. Recommendations have evolved from specifications comprising 10 ACH/hr, a positive pressure of 15 pascals relative to the corridor and high hygienic demand SUP (Supply air category) 1 filtration (National Services Scotland, 2011). ICUs housing immunosuppressed patients require HEPA filtration in the isolation rooms. Clean and dirty utility areas should be designed with 6 ACH/hr. The dirty utility should be extract only and at a negative pressure to the corridor with the clean utility at positive pressure to the corridor and supply air.

ICU’s should be at a positive pressure relative to the external corridor to protect against contamination, especially due to Aspergillus spp. Acinetobacter spp and MRSA (Baddley et al., 2013; Ichai et al., 2020).

ICU ventilation is classed as critical and following initial commissioning and validation should be subject to annual verification(National Services Scotland, 2011). HTM 0301 introduces the Ventilation Safety Group whose remit is the operational management and maintenance of ventilation systems in addition to annual verification and performance testing (NHS England, 2021b). Saran et al., advocated six monthly checks on indoor air quality quoting an acceptable Index of Microbial air contamination (IMA) of up to 25 (10–39 cfu/dm³/h) in ICU (Saran et al., 2020). This is based on the use of settle plates (1 hour) 1m from the floor and 1m away from walls/obstacles. The microbial count (cfu) is converted to an IMA value. For passive air sampling, five classes of IMA have been defined, representing an increasing level of contamination (Viani et al., 2020). Air quality checks are not current practice in UK units. The same review article found variation in international guidance regarding HVAC standards in ICU. With regards to filtration, only the Dubai Health Authority recommends the use of HEPA filtration in ICUs (Ichai et al., 2020).

Isolation rooms

HBN 04-02 specifies that no unit should have less than 20% of their beds as isolation rooms and ICUs in hospitals with neutropenic haematology patients may require up to 50%. Clinical input is essential in planning the number and type of isolation rooms. Final specification will depend on the patient population, demographics of the catchment area and any specialist services (IHFG, 2017). HBN 04–02 is vague when describing the ventilation requirements for isolation rooms. It specifies single rooms should be provided with a system that can provide source and protective isolation with lobbies present (NHS England, 2021a). It proposes a balanced supply and extract ventilation to each isolation room and gowning lobby with the lobby functioning as an airlock. It recommends the lobby should have a high and balanced supply and extract air change rate to be effective against airborne organisms. HTM 03–01 provides information on the specification for negative and positive pressure rooms with regards to ACH/hr, pressure differentials and filtration in an appendix table. Other features include self-closing entryway with adequate seal, sealed floors ceiling walls and windows and a monitoring system (Al-Benna, 2021; NHS England, 2021b).

CDC guidance contains more specification for different isolation facilities with anterooms to accommodate immunosuppressed patients, airborne infections and those with requirements for both protective and source isolation utilising pressurised anterooms (CDC, 2019). HBN04-01 Supplement 1 describes detailed specification for negative pressure room facilities and positive pressure ventilated lobby (PPVL) rooms but cites critical care units as an exclusion for PPVLs (NHS England, 2013). Despite this, PPVL rooms are listed in the HTM 0301-part A as an option for isolation facilities in critical care areas (NHS England, 2021b). This appears contradictory and can cause confusion. Poovelikunnel et al. found evidence of efficacy of PPVL rooms for protecting at-risk patients from airborne infection as well as source isolation of those with airborne infections. However, this study evaluated just two rooms over a period of eight weeks and it is not clear if these rooms were within an ICU (Poovelikunnel et al., 2020). The authors highlight concerns regarding the reliability of negative pressure rooms based on the findings of a study undertaken during SARS-1 in Hong Kong where it was found that 60 % of ensuite facilities were operating at positive pressure and > 90% of the corridor-anteroom or anteroom-patient room doors had a bidirectional flow (Li et al., 2007). In an assessment of rooms designed to be negative pressure in the US, it was found that the direction of airflow in the negative pressure rooms was not always correct. In the seven hospitals assessed none had routinely assessed the efficacy of negative pressure rooms, this points to issues with validation and maintenance. 52/115 (45%) designated negative pressure rooms had positive airflow to the corridor. High-risk areas including ICU and emergency rooms were not equipped to provide respiratory isolation (Fraser et al., 1993). Due to ICU rooms electrical and medical gas requirements, isolation rooms can be difficult to seal resulting in the inability to control airflow and placing the patient or staff at risk. Isolation rooms within ICU should be subject to regular maintenance in the same manner as the unit as a whole, including permeability tests (Bartley and Streifel, 2010).

Future ventilation considerations

General specification and isolation rooms

ICUs would benefit from more bespoke and detailed guidance discussing the basic specification but also incorporating advice and specification of isolation rooms and how these should be incorporated into the design. Some background to formulation of the guidance for design teams, detailing how the specification arose and the patient risks would be beneficial. At present, it is not clear to the reader why there is a need for the higher air change rates and positive pressure compared to general hospital wards.

Clarity is required on the proportion and types of isolation rooms to be installed in ICU settings, and the role if any for positive pressure ventilated lobbied (PPVL) rooms in this setting. An understanding of the patient population is essential to determine types of rooms required. Whilst PPVL rooms provide both source and protective isolation, there are exclusions in guidance to their use in critical care areas including airborne infections and severe immunosuppression. The current PPVL with ensuite constitutes a risk in terms of little used water outlets as such facilities are rarely used by the patient. It is worth considering whether the PPVL design could be modified and used to advantage in an ICU setting by allocating the ensuite space for waste disposal, completely separating waste and macerators from the patient room.

Any isolation facility in ICU must be subject to regular maintenance and undergo annual verification including permeability testing (NHS England, 2021b). As such, ventilation safety groups should include all ICU settings on their verification programme. For isolation facilities, there should be visual displays of pressure and alarms for pressure failure with advice for staff on appropriate action should failures occur. Staff education regarding the different types of side room ventilation facility available and appropriate patient placement, is advised, particularly in units where there are different types of isolation rooms. Signage on isolation room doors might be useful in this regard.

Given the risk to ICU patients from fungal infections including Aspergillus spp (7% of cases) and Mucor, HEPA filtration of the unit is worth considering for future design as the incidence is likely to be higher (Baddley et al., 2013; Machado, 2021). Aspergillosis as a complication of influenza is well described in ICU patients and has been reported in association with COVID-19 (Montrucchio et al., 2021). In a recent Italian study of 12 ICU’s, 12% of environmental samples (infusion pumps and patient tables) testing positive with A. fumigatus and the authors suggest surface sampling in preventing fungal infection as knowledge of fungi in the environment can guide prophylaxis treatment regimes (Prigitano et al., 2022). These findings support HEPA filtration as an additional prevention strategy and should be included in future guidance. Key considerations for ventilation in ICU are listed in Table 1.

Table 1.

Key ventilation considerations for ICU settings.

Involvement of clinical teams from outset of design and consideration given to patient population and any specialist units
Air changes of 10/hour
Unit positively pressurised to the corridor (+ 10 pascals)
Filtration – SUP 1 but consider HEPA for the entire unit
Number of isolation rooms, type and proportion of isolation facility required, for example, positive or negative pressure rooms required
Dirty utility at negative pressure, extract only and 6 ACH/hour
Clean utility, positive pressure, supply only and 6 ACH/hour
ICU unit as a whole and each isolation room should be listed as a critical ventilation system and following initial commissioning and validation should be subject to annual verification
Hospital ventilation safety group to be involved in annual verification and any planned upgrades/refurbishments
Education of staff on type of isolation rooms available, consider signage where there are mixed facilities
Installation of visual pressure indicators for isolation rooms and an alarm system to alert to pressure failure

Challenges of COVID and future planning for pandemics

The COVID-19 pandemic highlighted the importance of ventilation in NHS estates. Many hospitals are equipped to deal with new and emerging airborne threats in having some negative pressure rooms, to deal with small case numbers, however, once epidemic/pandemic levels are reached, they are quickly overwhelmed.

Challenges emerge for ICUs admitting both COVID positive and negative requiring ICU care. Larger facilities with multiple ICU/HDU units may be able to dedicate one or more for COVID-19 patients. In a South Korean hospital, one of two ICUs was remodelled due to lack of capacity for isolation of COVID-19 patients (Lee et al., 2020). The unit was divided into a space with 1) isolation room plus anteroom accommodating three beds and 2) two pre-existing airborne isolation rooms. Negative pressure was created in the anteroom by adding temporary duct systems which were connected to a pre-existing exhaust system. The isolation zone had five mobile negative pressure air machines generating a negative pressure relative to the anteroom. A negative pressure gradient was maintained between the existing airborne isolation rooms and the anteroom using an air volume control damper.

For smaller hospitals with one ICU, a conversion of this sort may not be possible and thought needs to be given to conversion of an HDU or general ward to a temporary ICU which may make it difficult to meet ventilation specifications (Peng et al., 2020). Miller et al. described conversion of a ward to a negative pressure facility (Miller et al., 2021). A temporary anteroom was set up in one hallway and the other was sealed off from the rest of the hospital by closing fire doors. Adjustments to the HVAC system were made to create negative pressure (−29 pascal) across the closed fire doors using two HEPA filtered negative air machines. This setup was achievable because the ward had its own dedicated AHU, bathroom exhaust system and a firewall separating it from the rest of the hospital. Negative pressure was sustained for a period of 24 hours. Whilst achievable it is not transferable to all settings and in a pandemic situation there is a need for prolonged and sustained negative pressure.

Conversion of ICU facilities to negative pressure throughout is not without risk. In one ICU hospital, previous air sampling had not shown any evidence of fungi and the annual incidence of Aspergillus cases was low (< 2%). Two months after converting rooms to negative pressure, 6 of 26 patients developed probable or proven pulmonary aspergillosis and air sampling from rooms of the first four patient cases recovered A. fumigatus (National Services Scotland, 2011). Increasing the pressure to 1.2 +/− 1.5pa resulted in a reduction in the recovery of the A. Fumigatus (0–2 CFU/m³) and the authors concluded that converting to negative pressure could have had unintended consequences and dispersed Aspergillus in dust from the false ceilings in the plenum spaces (National Services Scotland, 2011).

Consideration must be given to design of future ICUs with respect to pandemic preparedness. Units with 100% single rooms incorporating a PPVL design which would enable both protective and source isolation is an attractive option. Regardless, in a pandemic situation, surge capacity will also be required and consideration as to how this can be achieved without unintended consequences. A number of solutions for temporary isolation have been proposed. These include models that rely on air filtration/dilution, and those that rely on local exhaust ventilation (ASHE, 2020). These temporary facilities are unlikely to comply with guidance documents and may lead to contamination of clean areas and compromise fire safety (ASHE, 2020). As such, these proposals should be risk assessed by the ventilation safety group prior to implementation.

Conclusion

We have discussed ventilation aspects of ICU design including future preparedness for pandemics. Despite ICUs serving different patient populations, there are many design features that remain common to all, particularly with respect to ventilation systems and isolation rooms. Present guidance is piecemeal, and conflicting, and more bespoke guidance would be beneficial.

ORCID iD

Teresa Inkster https://orcid.org/0000-0003-1608-5224