A medical worker stands next to an oxygen cylinder at the Yatharth Hospital in Noida, on the outskirts of New Delhi, India, September 15, 2020. Photo: Reuters/Adnan Abidi
How widespread is the shortage of oxygen supply across Delhi’s hospitals? Does it vary by hospital type?
The Delhi government’s COVID-19 beds portal now provides information on oxygen beds at the city’s hospitals. It displays the most recent data provided by nodal officers at each hospital. There are currently around 180 hospitals in Delhi, across which nearly 20,000 oxygen beds have been allocated for COVID-19 patients.
The portal also provides crucial information on how long the oxygen supply at each hospital will last (at a given moment). For example, data on the morning of May 5, 2021 (figure 1), shows 102 hospitals had oxygen supply that would last for 24 hours or fewer. These hospitals accounted for nearly 70% of all oxygen beds.
Notably, the average occupancy of oxygen beds at these hospitals was remarkably high: between 90% and 100%. Of these, at 82 hospitals – which had about 55% of the total oxygen beds across the city – there wasn’t enough oxygen left for more than half a day (i.e. 12 hours or fewer).
The hospitals are of three types: Delhi government hospitals, Central government hospitals and private hospitals. On May 5, 14 Delhi government hospitals (Lok Nayak Jai Prakash Hospital, Guru Teg Bahadur Hospital, etc.) and 12 Central government hospitals (AIIMS, Ram Manohar Lohia Hospital, etc.) had allotted oxygen beds for COVID-19 patients.
The remaining oxygen beds were located in private hospitals, and the number of such beds for COVID-19 patients at these institutions varied by hospital size. Delhi’s largest private hospitals – the Sir Ganga Ram Hospital and the Indraprastha Apollo Hospital – had allotted 565 and 470 beds, respectively. At another 30 private hospitals, 100-400 beds had been allotted for this purpose. At the remaining institutions, fewer than 100 beds each had been allotted for COVID-19 patients.
Strikingly, the median number of hours for which hospitals had oxygen supply left (figure 2) was lowest among Delhi government hospitals – at 8 hours – and highest among Central government hospitals – at 14.5 hours. The difference in averages was starker. Delhi government hospitals had oxygen supply left for 11 hours and 4 minutes on average whereas Central government hospitals had oxygen for 22 hours and 24 minutes – more than twice as long.
The median number of oxygen supply-hours left at private hospitals was an intermediate 12 hours. Among private hospitals that had designated fewer than 50 oxygen beds for COVID-19 patients, the median oxygen supply remaining was for 10 hours. And for the remaining, comparatively larger, private hospitals, there was enough oxygen left for 12.5 hours.
However, among private hospitals that had oxygen left for 120 hours (five days) or fewer, the smaller ones had 8.5 hours of oxygen remaining and the larger ones, 12 hours. Again, the averages tell a similar story as in the previous case – as do the numbers from the mornings of May 6, 7, 8 and 9.
What explains the Central government hospitals’ higher oxygen supply? On May 5, there were 3,464 oxygen beds for COVID-19 patients at 12 Central government hospitals – whereas the 14 Delhi government hospitals had 4,388 beds. The median number of oxygen beds at Central government hospitals was 278.5, versus 173.5 at Delhi government hospitals. This is one explanation for the higher oxygen reserves at Central government hospitals.
Alternatively, the higher oxygen reserves at Central government hospitals might be the result of the international aid India recently received. According to a recent Union health ministry press release, of the eight hospitals and institutions in Delhi that have received medical equipment, including oxygen-related supplies (oxygen cylinders, concentrators, pressure swing adsorption plants and pulse oximeters), six are part of the Centre.
But one way or another, it also appears more oxygen may have been needed at the larger Delhi government facilities. Consider the association between the number of oxygen beds and the hours for which oxygen was left for at two types of government hospitals, as follows.
For government hospitals that had allotted 500 or fewer oxygen beds to COVID-19 patients – and for which data was available on May 5 – the data reveals a negative association for Delhi government hospitals and a positive one for Central government hospitals (figure 3). One implication is that the oxygen supply was more stressed at the larger Delhi government hospitals than at the larger Central government hospitals.
But on May 7 and 8, the association between the number of oxygen beds and the hours for which oxygen was left turned negative for both types of government hospitals. This suggests that the supply of oxygen was more stressed at the larger hospitals, potentially affecting more patients.
The data on the number of hours of oxygen supply left at Delhi’s hospitals also suggest that oxygen-related resources have been inequitably distributed among different types of hospitals that currently have oxygen beds for COVID-19 patients.
As multiple stakeholders in the government, the judiciary and civil society work to address the oxygen shortage at Delhi’s hospitals, it is imperative that they consider these patterns. Specifically, they could focus their efforts on hospitals where oxygen supply appears to be consistently low: the Ambedkar Nagar Hospital, the Deep Chand Bandhu Hospital, the Sanjay Gandhi Memorial Hospital, the Holy Family Hospital, the Indian Spinal Injury Centre, the Ayushman Hospital, N.K.S. Hospital and Madhukar Rainbow Children’s Hospital.
In the coming days, as more oxygen-related supplies arrive as international aid or from domestic manufacturers, they need to make their way to hospitals that have greater need for them.
The Delhi government released this data late into its COVID-19 outbreak, but as the outbreak continues, it can and should be used to distribute scarce resources better.
The author is very grateful for comments from Partha Mukhopadhyay and Shamindra Nath Roy at the Centre for Policy Research, New Delhi.
Shahana Sheikh is a PhD candidate at the Department of Political Science, Yale University. The views expressed here are her own.