New method for measuring compliance with long term oxygen treatment

BMJ 1994; 308 doi: (Published 11 June 1994) Cite this as: BMJ 1994;308:1544
  1. G D Philips,
  2. N K Harrison,
  3. A R C Cummin,
  4. J Ward,
  5. V S Shenoy,
  6. V Newey,
  7. R Ritchie,
  8. I P Williams,
  9. F J C Millard
  1. Chest Clinic, St George's Hospital, London SW17 0QT
  2. Department of Medical Physics Department of Chest Medicine, St Albans City Hospital, St Albans Al3 5PN
  1. Correspondence to: Dr Phillips
  • Accepted 7 February 1994

Long term oxygen treatment used for at least 15 hours a day improves survival in patients with chronic obstructive lung disease and respiratory failure.1,2 In 1989 prescriptions for oxygen for use at home, particularly oxygen concentrators, cost the NHS pounds sterling 18 million, and the number of such prescriptions has been increasing.3 There has never, however, been a completely reliable method for measuring compliance with domiciliary oxygen treatment. The most common method uses a clock incorporated into the oxygen concentrator, but patients may take off their nasal cannulas, and leaving the concentrator and clock running, or they may run the machine without wearing the delivery system.

We developed a method to assess compliance that uses plastic electrodes connected to a small (14.3x9.0x3.3 cm), battery operated timing device. This “electrode timer” records the length of time that the nasal cannulas are in contact with the skin (figure); it contains a non-volatile memory that enables recording for up to three weeks. The aims of the present study were to validate the electrode timer and to use it to assess compliance in patients with respiratory failure who were receiving long term domiciliary oxygen treatment with an oxygen concentrator and nasal cannulas.

Patients, methods, and results

In the validation study the electrode timer was used in 15 inpatients receiving oxygen treatment with nasal cannulas for acute infective exacerbations of chronic bronchitis or emphysema. The patients were monitored by nurses, who recorded how long the cannulas were worn but did not influence the positioning of the cannulas. A close correlation existed between the length of time recorded by the electrode timer and the length of time a patient received oxygen as observed by the nurses (r=0.996, P<0.001).

We then studied 56 patients who were receiving long term domiciliary oxygen treatment. All were visited in their homes, where plastic electrodes were cut to size, attached to the patients' nasal cannulas, and connected to the electrode timer. The patients were told to use their oxygen in the usual manner and that the device was to check the oxygen delivery system. One week later the timer was collected, and the times that were recorded by the concentrator's clock and by the electrode timer were noted. The study was approved by the ethics committees of the two hospitals participating in the study.

The mean (range) time for which patients were monitored was 162 (122- 193) hours. According to the concentrator's clock, the mean total time for which the concentrator was switched on was 124.7 (38.1-171.5) hours which suggested that oxygen was used for a mean of 18.2 hours a day. The mean total time, however, that patients received oxygen as determined by the electrode timer was 89.6 (27.3-162.2) hours, which suggested that oxygen was used for a mean of only 12.7 hours a day. This was a significantly shorter length of time than that recorded by the concentrator's clock (P<0.001), and no correlation existed between the time measurements of the clock and those of the electrode timer (r=0.36).


Our results suggest that the electrode timer is more accurate than a concentrator's clock for assessing patients' compliance with domiciliary oxygen treatment and that many patients seem to use oxygen for less than the recommended 15 hours a day.4 Previous studies that relied on the concentrator's clock to measure compliance may therefore have overestimated the daily oxygen requirement to reduce mortality in patients with respiratory failure. Our findings have potential economic implications and indicate that the daily oxygen requirement recommended to patients may need to be revised.


We thank the audit committees of St George's and St Albans Hospitals for financial support.


View Abstract