The hypothyroidism indicators are easy to achieve and pave the way to improving care for those at risk of the condition as well, explains Dr Matthew Lockyer


I last wrote about the clinical indicators for hypothyroidism in Guidelines in Practice, April 2004,1 just at the inception of the nGMS contract. I made two predictions, one of which was proved right and the other, discussed later on, was wrong.

I correctly predicted that most practices would view the hypothyroidism points as an easy target.These proved to be the easiest of all the clinical indicators to achieve with 98.2% of practices in England gaining full points for hypothyroidism.2 In contrast, the hardest to achieve were the COPD indicators, with only 86.2% of practices achieving maximum points.2

Thyroid indicator 1

Thyroid indicator 1 (2 points) requires a practice to produce a register of patients suffering from hypothyroidism (Table 1, below). The preferred Read codes remain CO4% (acquired hypothyroidism) and CO3% (congenital hypothyroidism).

Table 1: Clinical indicators for monitoring hypothyroidism
Disease/ indicator no Clinical indicator Points Qualifier Preferred Read code Exception reporting & Read codes Payment stages
Thyroid 1 Register of patients with hypothyroidism 2   Acquired hypothyroidism CO4%
Congenital hypothyroidism CO3%
Unsuitable/dissent N/A  
Thyroid 2 Percentage of patients with thyroid function tests 6 Recorded in the past 15 months Thyroid function tests 442% Patient unsuitable 9h71 Informed dissent 9h72 25-90%

This is one of the simplest registers to create using a practice computer because there is only one medication for the condition, thyroxine, and all the patients should be on it.

Thyroid indicator 2

Thyroid indicator 2 (6 points) rewards a practice for demonstrating that thyroid function tests have been performed within the past 15 months.

The points begin at 25% coverage and maximum points are achieved for 90% coverage.

The preferred Read code for thyroid function is 442%. Exemption is unusual but can be recorded for unsuitable patients, such as the housebound or terminally ill, or informed dissenters using codes 9h71 and 9h72, respectively.

Disease prevalence

The authors who published the consensus statement on hypothyroid management on which the current indicators are based also published a paper on the incidence of thyroid disease.3,4 They quote 3.5 per 1000 for women and 0.6 per 1000 for men, giving a female:male ratio of approximately 6:1.The paper also states that the probability of developing hypothyroidism increases with age.

In the past I found it hard to find a quoted figure for prevalence, even with this information. My incorrect prediction was that we would have a hypothyroidism population of 600 in our practice of 9000. This prediction was way out because we had carried out a badly constructed search that had counted patients twice (i.e. we had searched for patients diagnosed with hypothyroidism and patients on thyroxine) – an illustration of how easy it is to make this type of mistake.

At present we have 227 patients on our register. Our practice population is 9378. This gives a prevalence of 2.42%. In fact this is slightly in excess of the current national prevalence estimate of 2%, and will therefore marginally inflate our points reward for this clinical area.

Measuring thyroid status

Our local laboratory uses thyroid stimulating hormone (TSH) as a measurement tool for thyroid status, only measuring the T4 level if it is outside the normal range (normal TSH range 0.35-5.5 mU/l).The recommendation for ideal replacement therapy is a slightly suppressed TSH with T4 at the upper limit of normal.

TSH can sometimes be slightly outside the normal limits with intercurrent illness, but our structured checks produced a small number of patients with a significantly raised TSH and normal T4.

I discussed these with our local endocrinologist who was quite amused that I did not immediately realise the explanation.

These results are from non-compliant patients who only take their treatment when they are invited for a blood test. T4 levels respond quickly but the TSH is much slower to change. I had always assumed that because patients feel better on thyroxine there would be high compliance with treatment.


It is difficult to predict how the thyroid clinical indicators may evolve. Unlike other fields, such as diabetes and hypertension where best practice is continuously revised as new evidence is published and debated, hypothyroidism remains a quiet clinical backwater.Its consensus statement for management is unchanged for almost a decade.3

I continue to believe that it would be beneficial to extend the register to include those known to be at risk of developing hypothyroidism. Excluding patients with type 1 diabetes who should be offered thyroid screening as part of their package of annual checks, the list might include:

  • Patients with a previous history of hyperthyroidism
  • Patients treated with partial thyroidectomy or radioiodine
  • Patients taking amiodarone or lithium
  • Patients with a history of pituitary disease.

Extending the register in this manner would be in keeping with the consensus statement for the management of thyroid disorders from which the current indicators are taken. There is also evidence of the effectiveness of extending the register in general practice.5



Guidelines in Practice, December 2005, Volume 8(12)
© 2005 MGP Ltd
further information | subscribe

  1. Lockyer M. Register of patients with hypothyroidism is first step to improved care. Guidelines in Practice 2004; 7(4): 38-9.
  3. Vanderpump MP,Ahlquist JA,Franklyn JA, Clayton RN. Consensus statement for good practice and audit measures in the management of hypothyroidism and hyperthyroidism. Br Med J 1996; 313: 539-44.
  4. Vanderpump MP,Tunbridge WM,French JM et al. The incidence of thyroid disorders in the community: a twenty year follow up of the Whickham survey. Clin Endocrinol 1995; 43: 55- 68.
  5. Hill JP, Pitts-Tucker T. Consensus statement on management of hypothyroidism and hyperthyroidism. Registers based in general practice are essential in long-term surveillance. Br Med J 1996; 313: 1488 (letter).