Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Treating hip and knee arthritis in the UK

Study
Results
Whole initial cohort
Hip problems
Knee problems
Comment

Healthcare isn't just about interventions, but also about delivering care. It is about what happens in the real world, and if you want to make things better you really have to start with a clear picture of that real world. Asking experts, or coalface professionals, or both, is the way we usually get our snapshot. Because practice can vary, and numbers consulted are usually small, our snapshots can be unfocused and fuzzy. It is rare to have a high-definition picture, gained through detailed study, providing quality, size, and validity. One such provides a super picture of how hip and knee arthritis in older people is treated in the UK [1].

Study

Data were from the MediPlus primary care database in the UK, with about three million anonymised patient records. Practices with complete data up to 2002 formed the study group, about 60% of the total. The cohort was defined as people aged 65 or above who had a first consultation about a knee or hip condition in 1996 to 1998.

For inclusion there was a requirement of at least three years of records before and after the consultation, with problems in hip or knee (but not both together). Excluded were people with previous history of arthritis, chronic conditions likely to produce hip or knee problems, like rheumatoid arthritis, previous surgery, or recent evidence of trauma. Records were followed for three years from the consultation and information about tests, referrals, and therapy extracted.

Results

There were 311,000 people aged 65 years or older. In the cohort, there were 5,600 initial consultations for a hip problem, and 10,500 for a knee problem. When all the inclusion and exclusion criteria were applied, the final cohort had 1,410 people consulting for a hip problem, and 3,152 for a knee problem. Of these, about 65% were women, 35% were aged 75-84 years, and 7% aged 85 years or more.

Whole initial cohort

Table 1 shows a selection of results for treatments for the whole cohort during 1998, and for the patients with hip or knee in the year before the index consultation. Hip and knee patients had higher rates of referrals, and were prescribed more NSAIDs, particularly topical NSAIDs, and codeine or tramadol, but were no more likely to receive a COX-2 inhibitor. Antacid and anti-ulcer drugs were also more commonly prescribed.



Table 1: Referrals and medical treatments over one year in over-64s



Total or percent
Cohort
Hip
Knee
Number
310,843
1,410
3,152
Referrals of any kind
8.7
18.7
16.9
Physiotherapy
2.0
3.8
3.8
Safer NSAIDs
6.7
12.3
13.9
Other NSAIDs
2.3
5.9
5.1
NSAID/GPA combination
1.1
1.0
1.3
COX-2
0.6
0.6
0.6
Topical NSAID
4.6
8.5
10.1
Codeine, tramadol
20.8
37.6
36.9
Anti-gout
0.9
2.1
1.8
Antacids
5.3
12.5
12.8
H2-antagonist
3.9
6.5
8.3
PPI
3.9
6.1
6.1
Antidepressant
5.0
8.3
8.1
For cohort the data are for 1998, for hip and knee one year before index consultation


Hip problems

In the three years following an initial primary care consultation for a hip problem, about 40% of patients had been referred to medical or surgical consultation about the joint (Table 2), and one in 10 had had a joint replacement (Figure 1). About 1 in 12 of these patients had, by three years, also developed a knee problem (Figure 2).



Table 2: Cumulative percentage of patients with hip problems being referred or having medical treatments over three years



Cumulative hip patients
Cumulative percentage
First
12 months
36 months
Referrals of any kind
5.4
35.0
55.0
Referrals for joint
2.8
22.4
38.2
Physiotherapy
1.4
12.5
20.7
Safer NSAIDs
20.7
39.2
50.5
Other NSAIDs
4.6
12.2
16.8
NSAID/GPA combination
2.1
6.1
10.3
COX-2
1.4
3.6
7.7
Topical NSAID
6.1
18.0
28.8
Codeine, tramadol
31.6
65.9
79.4
Anti-gout
0.5
2.3
3.5
Antacids
1.6
14.4
22.4
H2-antagonist
1.9
9.9
16.4
PPI
1.1
8.1
17.7
Antidepressant
1.6
11.0
22.1




Figure 1: Cumulative percent having joint replacement surgery





Use of codeine analgesics, tramadol, safer oral NSAIDs and topical NSAIDs was common (Table 2), with other, less safe, oral NSAIDs, NSAID combined with gastroprotective agents, and COX-2 inhibitors less so. For these, for antacids and anti-ulcer drugs, and antidepressants, prescribing rates rose substantially over the three years after the initial consultation (Table 2).

Knee problems

In the three years following an initial primary care consultation for a knee problem, about 30% of patients had been referred to medical or surgical consultation about the joint (Table 3), and one in 50 had had a joint replacement (Figure 1). About 1 in 20 of these patients had by three years also developed a hip problem (Figure 2).



Table 3: Cumulative percentage of patients with knee problems being referred or having medical treatments over three years



Cumulative knee patients
Cumulative percentage
First
12 months
36 months
Referrals of any kind
4.2
30.4
48.8
Referrals for joint
1.3
17.8
31.5
Physiotherapy
2.4
11.1
17.7
Safer NSAIDs
26.2
41.4
51.9
Other NSAIDs
4.3
10.2
14.3
NSAID/GPA combination
2.4
5.3
8.1
COX-2
0.9
2.6
6.0
Topical NSAID
15.7
27.8
36.9
Codeine, tramadol
25.0
57.3
72.3
Anti-gout
0.5
2.3
3.5
Antacids
2.0
14.1
22.1
H2-antagonist
1.8
9.7
15.3
PPI
1.3
7.6
14.8
Antidepressant
1.3
9.9
18.5




Figure 2: Cumulative percent developing knee problem (with original hip problem, pale blue) or hip problem (original knee problem)





Prescribing of NSAIDs, oral and topical, codeine analgesics and tramadol, antacids and anti-ulcer drugs, and antidepressants all rose substantially over the three years (Table 3). Use of topical NSAIDs was higher in knee than hip patients, with use of codeine analgesics or tramadol rather less so.

In both groups there was also increased use of oral and injected corticosteroids (cumulative use about 7-10% of each for hip and knee). Use of morphine and similar analgesics was low, below 2%.

Comment

The results of this study are pertinent to the UK, and probably not for many other places in the world, because attitudes and practice patterns will differ markedly. Here more hip (10%) than knee patients (2%) underwent joint replacement within three years of first consultation. Elsewhere more patients may have received a joint replacement, and the difference between the proportions for knee and hip will be different.

But the methods used, and the size of the study, demonstrates the amount of information that can be derived from database studies, in order to help think about delivery of healthcare, and the capacity to deliver it. There's an old Irish joke about the tourist lost in rural Ireland who asks for help from a local, and gets the response: "If I wanted to go to Dublin I wouldn't start from here".

This study plonks a pin in the map from which one starts. It allows one to ask questions. In the discussion of this paper, for instance, the comment is made about adverse effects of medical treatments that carry high risks over the longer term. The data in this paper should be meat and drink for good health economic modelling of where we are now, and where we want to go.

Reference:

  1. L Linsell et al. Prospective study of elderly people comparing treatments following first primary care consultation for a symptomatic hip or knee. Family Practice 2004 21: doi 10.1093/fampra/cmh609.

previous or next story