Effectiveness of routine self monitoring of peak flow in patients with asthmaBMJ 1994; 308 doi: https://doi.org/10.1136/bmj.308.6928.564 (Published 26 February 1994) Cite this as: BMJ 1994;308:564
- N Drummond,
- M Abdalla,
- J A G Beattie,
- J K Buckingham,
- T Lindsay,
- L M Osman,
- S J Ross,
- A Roy-Chaudhury,
- I Russell,
- M Turner,
- J A R Friend,
- J S Legge,
- Health Services Research Unit, University of Aberdeen, Aberdeen AB9 2ZD
- Thoracic Medicine Unit, Aberdeen Royal Hospital, Trust, City Hospital, Aberdeen AB9 8AU
- Correspondence and requests for reprints to: Mr Drummond.
- Accepted 2 November 1993
Objective : To evaluate the effectiveness of routine self monitoring of peak flow for asthma outpatients.
Design : Pragmatic randomised trial.
Setting : Hospital outpatient clinics and general practices in north east Scotland.
Main outcome measures : Use of bronchodilators and inhaled and oral steroids; number of general practice consultations and hospital admissions for asthma; sleep disturbance and other restrictions on normal activity; psychological aspects of health including perceived control of asthma.
Results : After one year there were no significant differences between patients randomised between self monitoring of peak flow and conventional monitoring. However, those given a peak flow meter recorded an increase in general practice consultations that was nearly significant. Among patients whose asthma was judged on entry to be more severe, those allocated to self monitoring used more than twice as many oral steroids (2.2; 95% confidence interval 1.1 to 4.6). Patients who already possessed a peak flow meter at the start of the study recorded higher morbidity over the course of the year than those eligible for randomisation.
Conclusion : Prescribing peak flow meters and giving self management guidelines to all asthma patients is unlikely to improve mortality or morbidity.Patients whose asthma is severe may benefit from such an intervention.
Peak flow meters have become ubiquitous in the diagnosis and management of asthma in general practice
The effectiveness of peak flow meters as self monitoring devices remains in doubt
Patients randomised to receive a peak flow meter and self management guidelines experienced little clinical benefit over the course of one year
Peak flow meters as self monitoring devices should be targeted at patients whose asthma is more severe or difficult to treat
In October 1990 mini peak flow meters were made available on prescription in the United Kingdom, in response to several years of campaigning by the medical profession and the National Asthma Campaign for the wider use of what was regarded as a crucial instrument in asthma control. The device has been thoroughly validated as an accurate measure of peak expiratory flow rate,1 which is highly correlated with other measures of pulmonary function.2 Predictive values for flow rate have been calculated.3 Furthermore, the device is inexpensive, simple to use, and easy for patients to understand.4 While spirometry remains the usual method of assessing pulmonary function in hospital practice, peak flow meters have become widely used in the management of patients by general practitioners.5 Many patients are being given peak flow meters to take home, or are acquiring them themselves, and the British Thoracic Society has recommended the use of home recordings for good asthma management.6
There have, however, been few controlled studies of the effectiveness of mini peak flow meters in self management of asthma. Beasley et al issued peak flow meters and self management guidelines to 36 patients and recorded significant improvements in objective and subjective measures of asthma severity over six months, but their study lacked a control group.7 Janson-Bjerklie and Schnell, in a controlled design, issued meters to 15 patients, who reported decreased use of prescribed bronchodilators over two to three months, but no difference in symptoms.8 Charlton et al compared peak flow and self management plans based on symptoms and found no significant difference between patients.9 They concluded that issuing patients with peak flow meters alone would not improve asthma control.
The Grampian asthma study of integrated care was designed to evaluate the effectiveness of self monitoring of peak flow, integrated care, and personalised, computer based education for asthma patients. This paper reports our evaluation of peak flow self monitoring. The evaluations of integrated care and enhanced education are reported separately.10,11
The Grampian asthma study of integrated care was constructed as a 2x2x2 randomised trial1,2: patients were independently assigned at random between self monitoring and conventional monitoring of peak flow; between integrated and conventional care; and between enhanced education and conventional education. Patients who already owned a peak flow meter were excluded from randomisation for that dimension of the study but were considered for randomisation on the two remaining dimensions.
Patients who were over 16 years, had their diagnosis of asthma confirmed by a specialist, had pulmonary function reversibility of 20% on treatment, were attending outpatient chest clinics in Aberdeen, Banff, Elgin, and Peterhead, and were not already inpossession of a peak flow meter were allocated at random to either self monitoring or conventional monitoring of peak flow. If allocated to self monitoring, they were shown how to use a mini-Wright peak flow meter and given a device to take home, along with self management guidelines to follow when they identified changes in peak flow. Each set of guidelines was tailored to the patient's history of asthma according to the clinical judgment of the recruiting physician. Typically, the guidelines included a level of peak flow rate below which the patient should start a course of oral steroids, and a level below which the patient should seek immediate medical assistance. Patients allocated to conventional monitoring were given no self management advice beyond that normally offered during an outpatient consultation. Each patient participated in the study for one year.
All patients were reviewed every three months, by their general practitioner or by a hospital doctor. Before each review, self monitoring patients were sent a peak flow diary card and asked to complete it over the week before visiting the doctor. At each scheduled consultation during the study year, the doctor was asked to record relevant information, including pulmonary function, number of prescriptions for asthma drugs, amount of sleep disturbance caused by asthma in the previous week, and extent of restricted activity in the previous month. Full details of the methods of randomisation, data collection, and statistical analysis are described elsewhere.10 When peak flow meters became available on prescription, all general practitioners in Grampian were reminded of the nature of the study in order to minimise contamination through patients randomised to receive conventional monitoring but later being prescribed a meter.
A total of 801 patients gave informed consent and were randomised between the three innovations. A total of 232 patients were already in possession of a peak flow meter and were therefore ineligible forrandomisation for that dimension of the study. Nevertheless patients who took part in the randomised trial were similar on entry to those excluded for their pulmonary function, perceived anxiety, self efficacy, and asthma control scores. The exception concerned social and physical functioning: patients eligible for randomisation recorded a mean score on the living with asthma scale that was higher by 0.23 (95% confidence interval 0.11 to 0.35). Though small in clinical and social terms, this difference suggests that patients who participated in the trial had less severe asthma on entry to the study than those who already owned peak flow meters.
In all, 569 patients were randomised, 285 to self monitoring and 284 to conventional monitoring, thus providing 80% power of detecting at the 5% significance level a difference between peak flow monitoring and conventional monitoring equivalent to 23% of the standard deviation of the variable in question. The initial interviews confirmed that randomisation had been effective: table I shows patient characteristics at the beginning of the study.
After 12 months pulmonary function did not differ significantly between the two randomised groups, as assessed by forced expiratory volume in one second (as percentage of predicted) and by peak expiratory flow rate (table II).
Table III summarises the effect of self monitoring on the five clinical measurements, after the patient's initial peak flow, forced expiratory volume in one second (as percentage of predicted), and duration of asthma were controlled for. For use of bronchodilators, oral steroids, and inhaled steroids and the number of general practitioner consultations and hospital admissions during the year, the analysis showed no significant differences between the two randomised groups. However, the number of consultations for the self monitoring group approached statistical significance, with an estimated mean increase of 20% (0% to 45%).
Within the randomised groups, 47% of the patients (226/485) never reported sleep disturbance (48% (114/239) in the peak flow meter group and 46% (112/246) in the control group) and 6% (29/485) reported that their sleep was disturbed every nightb (14/239 in the peak flow meter group and 15/246 in the control group). Analysis confirmed that type of monitoring was not significantly associated with this distribution of sleep disturbance. Table IV shows the mean number of disturbed nights in a week for those patients who sometimes reported sleep disturbance. No significant differences were found between the two randomised groups. Similarly 73% of patients (334/455) reported no restriction of activity resulting from asthma (76% (166/220) in the peak flow meter group and 72% (168/235) in the control group). This distribution was not significantly associated with randomisation to self monitoring or not. For those patients reporting some restriction of activity, the mean number of restricted days in a month is shown in table IV. No significant differences were found between the two randomised groups.
After 12 months, 538 (94%) of those randomised and 219 (94%) of those who already had a peak flow meter were interviewed in their homes. Response rate was similar in the two randomised groups. There were no significant differences between patients in the two randomised groups for any of the psychological outcomes (table IV). A comparison of the group who already owned a peak flow meter and the randomised groups of patients is also given in tables II-V. The results confirm that those patients who already owned a peak flow meter at the start of the study suffered from more severe asthma.
Interaction between type of monitoring, integrated care, and enhanced education
Tests of interaction between the type of monitoring and any combination of the other two innovations (including patients whose asthma was too severe for them to be randomised to integrated care) were carried out. Table VI summarises only those variables for which type of monitoring had a significant interaction effect. After 12 months in the study, the subgroup of patients with asthma judged “too severe” and maintained in conventional care showed a significantly higher use of oral steroids if they had a peak flow meter (ratio of the means 2.22 (1.08 to 4.59)). Integrated care patients who reported variation in their limited activity had a significantly lower mean number of days of limited activity if they had a peak flow meter (ratio of the means 0.41 (0.22 to 0.75).
This study evaluated the effectiveness of miniature peak flow meters for hospital outpatients in a large randomised trial over 12 months. The results are therefore relevant to the debate about the circumstances under which meters should be prescribed to patients diagnosed as having asthma.
We found no differences in the pulmonary function of patients after 12 months. However, those patients ineligible for the randomised trial because they already owned peak flow meters required much more medical intervention during the year to maintain their pulmonary function, implying that their asthma was either more severe or more difficult to treat than those eligible for randomisation. These patients may benefit considerably from owning a meter, but this hypothesis remains untested in a prospective, controlled trial. Our findings relating specifically to patients eligible for randomisation suggest that peak flow meters, even when accompanied by guidelines for self management neither reduce morbidity nor improve anxiety, depression, self efficacy, or social and physical functioning. However, our findings do suggest that possessing a meter leads to some improvement in management for the more serious patients, in the form of increased use of oral steroids. Although only marginally significant, the tendency towards more general practice consultations among those in possession of a meter could also be important. Evidence suggests that many avoidable deaths from asthma are caused by delay in seeking medical help.22 Avery et al found that 68% of asthmatic patients were not seeing a doctor regularly.23 Guidelines that advise patients to seek medical help at a specified peak flow rate relieve them of a difficult decision and may result in earlier treatment of an acute episode. However, this study showed no difference in the hospital admission rate between those randomised to receive a peak flow meter and those who did not receive a meter.
These results contrast with those achieved by Charlton and colleagues, who reported reductions in general practitioner consultations and in oral steroid consumption for all patients, irrespective of the nature of their management plan.9 Theirs was a sample of patients derived from a single general practice rather than an outpatient clinic, however, and all patients in their study were subject to review by an asthma nurse at least once every eight weeks. Such intensive review may not be acceptable to patients over an extended period of time and is not feasible for many general practitioners.
Without objective evidence in the form of reductions in hospital admissions or improvements in pulmonary function, the benefits of a peak flow self management plan are questionable. In particular, it is not clear whether more or fewer consultations should be seen as beneficial. Charlton et al considered the reduction in consultations that they reported as evidence of better health,9 but we judge that the increase in general practice consultations among our patients represents better monitoring by health professionals.
This study reinforces the mounting evidence that self monitoring of peak flow does not improve morbidity. No study has been able to repeat the improvements reported by Beasley et al, who reviewed their sample of patients three times over two months and again after six months.7 This close monitoring of patients in hospital clinics may have substantially improved patient compliance with treatment. Our study showed only marginal improvements for patients given a peak flow meter and management guidelines and then followed up quarterly over a year. Perhaps disappointingly, our analysis of interaction showed that this remained true for patients if they were, in addition, receiving educational materials. The suspicion remains, however, that those patients most likely to benefit from owning a peak flow meter because their asthma was more severe or unstable were ineligible for our study because they already owned one.
We therefore conclude that issuing peak flow meters and self management guidelines to all asthma outpatients is unlikely to result in general improvements in morbidity. However, self monitoring may facilitate monitoring by general practitioners and encourage consumption of oral steroids by those whose asthma is more severe or unpredictable. We suggest that peak flow meters be targeted at these patients.
We thank all the patients and general practitioners who contributed to this study. We are indebted to Drs Aileen Flett, Geoffrey Packe, and Stephen Watt for clinical support; to Annette Kinsella, John Lemon, Simon Naji, and Jim Petrie for methodological support; and to Jennifer Birdsall, Alison Ewen, Jackie Fiddes, and Rae Glass for secretarial support. The study was funded by a project grant from the Medical Research Council; the core of the Health Services Research Unit is funded by the Chief Scientist Office of the Scottish Home and Health Department; neither of the funding bodies is responsible for the opinions expressed in this paper.