A national landscaping survey of critical care services in hospitals accredited for training in a lower-middle income country: Pakistan

Access and organisational structure

Average beds within teaching institutions per 100,000 population was 0.7 (total number of CCU beds in participating institutions divided by the total population of Pakistan), ranging from 0 in Gilgit Baltistan to 1.0 in Sindh. The median number of critical care beds per unit was 9 (range 0-30). A total of 58 (47.9%) CCUs were managed directly by the government and 42 (34.7%) of units were managed by the private sector, with the remainder being administered by not-for-profit organisations (17.4%). Seventeen (10.8%) of the CCUs surveyed reported a ‘closed’ model of care whereby intensivists were the consultant in charge of care. Unrestricted visiting for families was practiced in 40 (25.5%) of units (Table 3).

Infrastructure

Ventilator to bed ratio of 1:1 was observed in 82 (52.2%) of the CCUs (table 4), with Punjab province having the greatest number 49 (59.8%) and Azad Jamu Kashmir the lowest 0. All CCUs had a telephone line, however, only 52.2% had access to the internet. Table 5 summarises the availability of equipment to monitor critically ill patients. Almost all CCUs (95.5%) had access to 1:1 non-invasive multiparameter monitoring. Invasive arterial monitoring and capnography was available in 69 (44.0%) and 9 (5.7%) CCUs respectively. In addition, 51 (32.5%) and 39 (24.8%) units had access to point-of-care haemoglobin and lactate measurement respectively. Isolation rooms essential for management of infectious diseases, including severe acute respiratory infections, were available in 45 (28.7%) CCUs.

Human resources, team structure and training opportunities

Table 6 summarises the human resources, team structure and training opportunities. A senior clinician trained in critical care was available in only 19 (12.1%) of units. This is a ratio of one trained intensivist for every 82 CCU beds and 0.009 trained intensivists per 100,000 population. The majority of units (86, 54.8) was overseen by a consultant anaesthetist. In the remainder, 38.8% were overseen by a consultant physician and 6.4% by a consultant surgeon. A non-consultant doctor was assigned to CCU round-the-clock with no other work commitments in 140 (89.1%) of CCUs. Of the 121 institutions surveyed, 101 (83.5%) were recognised by the College of Physicians and Surgeons for speciality training (residency training). Critical care medicine (CCM) fellowship training was offered by 6 (4.9%) of institutions.

The majority of critical care units were managed by registered nurses with general training 149 (94.9%), with the remaining 8 (5.1%) being managed by technicians trained in anaesthesia or critical care. One to one nurse to bed ratio during the day for ventilated patients was available in 84 (53.5%) of units, and in 19 (12.1%) of units for self-ventilated patients. At night this availability dropped to 75 (47.8%) and 14 (8.9%) respectively. Availability of 1:1 nursing also varied between provinces, ranging from 56.5% having a 1:1 availability in the Punjab compared to 0% in Azad Jamu Kashmir. Microbiologists and haematologists were accessible in 96 (61.1%) and 103 (65.6%) of units respectively. Health care assistants or trained technicians were part of the care provision team in 94 (59.9%) of CCUs. Radiology technicians were available in 150 (95.5%) units and a further 128 (81.5%) CCUs had access to physiotherapy services.

Critical care capacity

The number of critical care beds in LMICs are known to be lower when compared to high-income countries [7,8], this disparity is pronounced in Pakistan in comparison to neighbouring countries: At 0.71 per 100 000 population, it is lower than Sri Lanka (2.3 critical care beds per 100,000), Nepal (2.8) and India (2.3) [9]. The survey further identified a wide disparity in access to critical care beds between the provinces (Figure 3). Punjab, whilst being the most densely populated province of the country, has lower availability of critical care beds than neighbouring Sindh. Similar disparity exists between major cities in each province (Table 2). As urbanisation and migration to cities for employment continues in Pakistan, and as the burden of non-communicable disease rises - including road traffic accidents and multimorbidities, it is a national priority to address the disparity in access to critical care services [2].

CCUs in both public and private sector institutions (including not-for-profit) had the basic infrastructure (electricity and a backup generator, piped oxygen, medical air and suction, infusion and syringe pumps), and basic monitoring (non-invasive multiparameter monitor, mercury thermometer, and manual CVP measurement). Overall ventilator to bed ratio was 1:1.3, meaning 3 out of every 4 CCU beds have the facility to mechanically ventilate. Availability of these resources is reassuring, and suggests that the provision of the mainstays of critical care organ support- ventilation therapy, basic cardiovascular monitoring and support, and delivery of fluids is possible. The safe and effective delivery of these therapies, however, relies not only on the availability of equipment, but on specialist trained staff with the skills to instigate, titrate and troubleshoot treatment.

In contrast to the specific resources of critical care described above, sinks for hand washing were absent in 29.3% of CCUs and access to isolation rooms or cubicles to control cross infection with negative/positive air exchange mechanism was available in just 28.7% of CCUs, the majority of which were private sector tertiary care hospitals in the major cities. Addressing the absence of facilities for infection control is perhaps a key priority for those seeking to improve critical care services in the country, given the increasingly important role critical care plays in the preparation and management of seasonal epidemics (including severe acute respiratory infections-SARI) and in the rising burden of drug resistant infections. Furthermore, access to point-of-care (POC) measurements including lactate haemoglobin, and availability of invasive haemodynamic monitoring, which are increasingly seen as essential resources for the management of critical illness, is lacking [10,11]. Only 39 (24.8%) of units had access to POC lactate and just 14 (8.9%) could invasively monitor haemodynamics. As Pakistan seeks to improve diagnosis and management of critically ill patients with SARI, sepsis and following trauma, better access to POC services and invasive monitoring, along with specially trained staff to interpret and respond to this information, is essential.

Capacity for training in critical care

The ratio of trained intensivists for each CCU bed in this survey is one to 82, much lower than estimates from South Asia, Latin and North America [12–14]. Critical care has been a recognised speciality in Pakistan with a structured training programme since 2004, however, at the time of this survey, only six institutions out of 121 teaching institutions (excluding military sites) were recognized by the College of Physicians and Surgeons of Pakistan for Critical Care Medicine training [15]. To date, just twenty-seven fellows have obtained the fellowship from the College in Critical Care Medicine [16]. Many currently practising intensivists in CCUs still have to travel outside of Pakistan for their higher training fellowships. Low intensivist to patient ratios (< 1:14) in academic medical CCUs have been cited as a barrier to delivery of quality of care and having a detrimental effect on staff well-being, specifically to the quality of pastoral and professional mentorship available for rotating trainees, who may consider specialising in CCM [17]. Lack of specialist training opportunities may be perpetuating the low numbers of designated critical care doctors on-call in CCUs and the low percentage of CCUs which are led by a trained intensivist. Lack of training opportunities extends beyond doctors, with only 40.1% of nurses in charge of CCUs having received any formal training in intensive, critical or cardiac care. Access to microbiologists, specialists who are increasingly considered fundamental to the interdisciplinary management of critically ill patients, was limited (61.1%).

Whilst there is growing evidence to support that intensivist-led patient management is associated with better patient outcomes and greater compliance with broadly accepted indicators of critical care quality [18], a closed model was uncommon in the CCUs surveyed (10.8%). A further 25.5% of CCUs reported a semi-closed structure, but without a trained intensivist as their clinical lead. Investment in the reorganisation of critical care services to improve operational efficiency and patient outcomes (length of stay, duration of mechanical ventilation) in CCUs has resulted in a shift toward closed organisational structures whereby admission to and management of patients within the CCU is coordinated by designated critical care clinicians [18]. Given the paucity of critical care services in Pakistan, such a model of management may promote effective resource utilisation. However, such models require national level investment in specialist training, and institution level investment in hiring and retaining such a specialist workforce. Other settings have demonstrated how investment in critical care trained clinician staff to lead units and investment in training for nurses working in critical care has positively impacted on safety within CCUs and outcomes for critical care patients. Without this investment, efforts to strengthen specialist capacity and improve quality of critical care services through research and implementation will be hindered.

Limitations

This survey only approached institutions recognised for specialist teaching. Consequently. the number of CCU beds per 100,000 population is underestimated. A recent multicountry snapshot of critical care bed availability [9], to which Pakistan contributed, reported a national average in Pakistan of 1.3 beds per 100,000 population. However, such estimates included units which may have no recognised affiliation with critical care training and no support from critical care societies. Whether considering the numbers reported here, or the estimates from those with broader inclusion, CCU bed availability and trained, skilled staff is still lower than neighbouring countries. Pakistan has no central register or standard definition for CCUs and as such units may be operating without the support of trained intensivists.