Guidelines in Practice Award runner up Dr Tom Fryatt explains how his practice responded to the challenge presented by the RCP’s glucocorticoid-induced osteoporosis guidelines


Glucocorticoids are commonly prescribed in general practice; oral glucocorticoids are associated with a significant increase in risk of fracture at the hip and spine.1

In Trent in 1996, 0.5% of the total population were found to be taking steroids for at least 3 months. Of these, only 14% were receiving prophylactic treatment to prevent glucocorticoid-induced osteoporosis.2

Guidelines developed jointly by the Royal College of Physicians, the Bone and Tooth Society and the National Osteoporosis Society on glucocorticoid-induced osteoporosis and published in December 2002 presented our practice with an opportunity to identify patients at risk and monitor our management of the prevention of glucocorticoid-induced osteoporosis.1

We had previously audited our performance in 1999 using local guidelines developed from the National Osteoporosis Society guidelines published in 1998. These guidelines recommended prophylactic treatment for patients receiving more than 7.5 mg prednisolone or equivalent daily. We had concluded that identifying patients being prescribed more than 7.5 mg of prednisolone was beyond our technical capability.

The 2002 RCP guidelines advise carrying out an assessment when treatment with glucocorticoids is expected to last 3 months or more, regardless of dose. This has enabled us to use our computer prescribing to audit and monitor our performance in providing prophylactic therapy to patients taking steroids over the long term.

Practice situation

Our practice, which is situated in a suburban area, has a population of slightly more than 10 800. There are six PMS principals, one nurse practitioner and five practice nurses.

The practice has been computerised for more than 10 years and is currently using Torex System 6000. All acute and chronic prescriptions are issued using the computer. Prescriptions issued on home visits are subsequently recorded on the computer system.

The practice has a policy of issuing repeat prescriptions for 50 days’ treatment.

How we implemented the guidelines

The RCP guidelines were discussed at a practice education meeting in October 2003. The concise guide version of the RCP guidelines, which contains a management algorithm, was distributed to each consulting room.

We first identified those patients who had been taking glucocorticoids for at least 3 months (Box 1, below). Using the report software supplied with the computer system we were able to identify patients whose records showed that they had received a prescription for glucocorticoids in the previous 50 days. This group was searched further to identify those with a prescription issued during the 50 days before that.

Box 1: Summary of search criteria
  • Select patients with glucocorticoid prescription during past 50 days
  • Select those with glucocorticoid prescription during previous 50 days
  • Exclude patients taking concurrent prophylactic treatment
  • Exclude patients having had DXA scan during past 3 years
  • Remaining patients form at-risk group
  • At-risk group records scrutinised
  • Repeat every 50 days

We then searched this group to exclude those who had a concurrent prescription for treatment to prevent osteoporosis (bisphosphonates or hormone replacement therapy – HRT).

The guidelines recommend that patients under 65 years treated with glucocorticoids should be assessed for fracture risk using dual energy X-ray absorptiometry (DXA) to assess bone mineral density. Therefore, we carried out a further search to identify patients whose records showed that they had undergone a DXA scan during the previous 3 years.

The remaining patients were those at risk of glucocorticoid-induced osteoporosis, who were not taking prophylaxis or had not undergone DXA scanning.

A report of this type is run every 50 days by practice clerical staff. The list of patients generated by computer is reviewed by one of the partners who then examines the computer record against the audit standards (Box 2, below), and alerts the prescriber if appropriate.

Box 2: Audit standards
  • Patients being prescribed glucocorticoids for at least 3 months should be managed according to the RCP guidelines


  • Patients over 65 years of age identified should be considered for bisphosphonate therapy

  • Patients under 65 years of age should be considered for DXA scan or treatment if they have a previous fragility fracture


The prescriber responsible for each patient decides on the appropriate intervention. If the identity of the prescriber is not clear from the record, the medication authorisation is cancelled, ensuring that the patient attends for review before further medication is issued. We did not consider it necessary to put an alert in the patient record at this stage.

Initial results

The results of an initial audit in October 2003 found 31 patients, 18 of whom were women, who had been given two prescriptions for glucocorticoids in the previous 3 months. This represented 0.25% of our practice population. Of these, we identified 11 men and nine women who were not receiving prophylaxis.

Table 1 (below) shows those who were receiving prophylactic treatment.

Table 1: Comparison of patients taking glucocorticoids and receiving treatment to prevent osteoporosis, at baseline and 6 months later
  October 2003 April 2004
  Women (<65) Men (<65) Women (<65) Men (<65)
Total taking glucocorticoids 18 (4) 13 (2) 26 (3) 14 (5)
Taking bisphosphonates 8 (0) 2 (1) 15 (2) 8 (2)
Taking HRT 1 (1) - 1 (1) -
DXA scan in previous 3 years 3 (1) 4 (0) 6 (1) 3 (0)
% treated 50 (25) 15.4 (50) 61.5 (100) 57.2 (40)

Second audit

In April 2004, we identified 26 women and 14 men in the at-risk group, and of these, 6 men and 10 women were not receiving prophylactic therapy. However, as Table 1 shows, the percentage of patients taking oral glucocorticoids who were receiving prophylactic treatment had increased markedly.

Reasons for non-compliance

Examination of the records of the individuals identified by the search sheds some light on why compliance with guidelines is not complete.

In the under-65 age group we identified two men who were receiving steroids following transplantation and a third who was receiving palliative care. In the over-65 age group, two men had recently had normal DXA scans and two very elderly women had significant concomitant disease.

How successful was the project?

We found that identifying patients using the software supplied with the computer system worked well. No additional work by partners or staff was required.

However, we discovered that there are a number of reasons why problems can arise in identifying patients correctly:

  • The practice prescribing policy is ignored or circumvented.
  • Patients occasionally ‘stock up’ on medication, especially before a holiday, so they may have more than two prescriptions in a 50-day period and none in the following 50 days.
  • Medication issued by hospital out-patient departments is not currently recorded on our practice computer system.
  • Poor compliance in taking bisphosphonates is being increasingly reported. Our system allows us to identify those not filling their repeat medication prescription and who may therefore not be compliant.

DXA scanning

Before December 2002, we recorded DXA scans in some patients aged 65 years and over, against the guidelines’ recommendation. However, we searched records from the previous 3 years as the guidelines recommend this period for monitoring those with T scores between 0 and ­1. There is open access to DXA locally although the waiting time for a scan is around 4-5 months.

Benefits of the project

Undertaking the audit has brought the following benefits:

  • An increase in concurrent prescribing of bisphosphonates
  • We have been able to identify poor compliance with prophylactic medication
  • We have increased awareness of the guidelines among practice staff
  • The scheme has proved a useful aid to education.

The future

We have deliberately kept this system simple to start with to enable us to assess the size of the problem and avoid over-reliance on technology. However, we now feel confident enough to add some refinements.

We plan to add an exemption code to the records of those for whom treatment is inappropriate.

We will also be adding a search for calcium and vitamin D prescriptions, as this is increasingly recommended as an adjunct to treatment.

Our practice pharmacist is keen to help monitor compliance.

The report we generate has been developed for use with the EMIS system as well as Torex 6000, systems which are used by 16 of the 19 practices that form Greater Derby PCT.

The report system has been modified to produce anonymous data, which also contribute to a PCT-wide audit of osteoporosis. Practices participating have been given the results, enabling each one to view its own performance in comparison with that of others.

A neighbouring PCT is considering implementing a similar system, and we are currently advising them on it.


  1. Royal College of Physicians, the Bone and Tooth Society, the National Osteoporosis Society. Glucocorticoid-induced osteoporosis: Guidelines for prevention and treatment. London: RCP, 2002.
  2. Walsh LJ,Wong CA, Pringle M,Tattersfield AE. Use of oral corticosteroids in the community and the prevention of osteoporosis: a cross sectional study. Br Med J 1996; 313: 344-6.

Guidelines in Practice, December 2004, Volume 7(12)
© 2004 MGP Ltd
further information | subscribe