Trevor Silver Prize Winner 2013

My Hip, My Knee; Who Should I See?

Osteoarthritis – A Condition For Primary Or Secondary Care?

Word Count - 2568.
Dr Caroline Guy, MBChB.
GPST 1.1 Paisley and Renfrewshire.

Musculoskeletal problems such as osteoarthritis make up a significant proportion of general practice consultations, up to 15% of the workload (1).  Knowledge and experience in this area is invaluable as a qualified general practioner (GP).   Despite this only 10% of vocational GP training schemes in the UK include an orthopaedic rotation (2).    I identified orthopaedics as a specialty in which I had little experience therefore opted to carry out a six month rotation in the Orthopaedic Department of Inverclyde Royal Hospital in Greenock.  I attended various outpatient clinics, trauma on call shifts and elective surgery where I was able to increase my knowledge and confidence in managing a wide range of orthopaedic conditions.  I was also able to see how referrals from the primary care are made and the expectations that Specialists in Orthopaedics have towards the general practioner (GP). 

The Consultants were concerned that in some cases referral letters from GPs provided little information regarding the patient, their symptoms and their history, but also that some patients had not received conservative management for their condition before referral.  The referral letter is the main means of communication at the interface between primary and secondary care (3).  Good referral letters have a positive impact on the efficiency and resourcefulness of patient care while a poor one gives little background history of the patient and may in some cases lead to urgent cases being missed or duplicate investigations being carried out (4).  I decided to review the quality of GP referral letters to the Orthopaedic Department in Inverclyde Royal Hospital, the appropriateness of the referral and the outcomes from the consultation.  To do this I reviewed 100 referral letters to secondary care at random across the 6 consultant led clinics.  A proforma was made to review each letter (see appendix 1). 

The orthopaedic diagnosis in the consultations from the cohort covered a spectrum of orthopaedics conditions, in this essay I will specifically lay additional focus on osteoarthritis: what the outcomes of these patients were and how we could improve our current practice.   I was also able to compare the management of these patients with the current NICE recommendations on osteoarthritis and on reflection of this, several interesting discussion points arose.

The first question the Consultant was asked was the patients’ diagnosis following clinic appointment.  Osteoarthritis (OA) was the most common diagnosis seen in the 100 randomly selected GP referral letters at 21%.  The second most common problem was back pain either mechanical or nerve root impingement (20%), and third was rotator cuff injuries (9%).

The table below shows the outcomes from the patients diagnosed as having osteoarthritis at clinic.

Outcome from Outpatient Visit

Number of patients / Percentage of OA


1 (4.7%)

Physiotherapy and analgesia

1 (4.7%)


2 (9.5%)


2 (9.5%)


2 (9.5%)

Patient declined intervention

3 (14.2%)

No treatment required

5 (23.8%)


5 (23.8%)

NICE outline core conservative management options that should be carried out prior to referral.  This includes education and exercise for all patients with osteoarthritis, with the addition of weight reduction for patients with a high BMI (5).  Further management options would then include paracetemol and non steroidal anti-inflammatory drugs (NSAIDs), as well as manual therapy, gait aids and orthotic devices.  If a patients’ symptoms persist despite these conservative measures, or their symptoms are interfering with their daily activities or sleep; referral to specialist care is recommended for consideration for arthroplasty (5).  As a consequence, some professionals may consider the ideal outcome from an orthopaedic clinic appointment to be surgical intervention for patients with OA but in this audit only 5 (24%) patients with OA went on to be listed for arthroplasty.  Thirteen (62%) of these patients received conservative management or no treatment at all. 

The second question the Consultant was asked to complete was the appropriateness of referral.  This was a subjective question and bias may have contributed to their answer.  However there was a box provided for justification of the decision.  Nineteen percent of all referrals were deemed inappropriate.  Three of these were patients with a diagnosis of osteoarthritis (14% of all patients with OA).  The Doctors justification for this included - early OA or had not yet received the core conservative measures for their disease.  In these cases the outcome from the consultation was physiotherapy, NSAIDs or analgesia.  The next task for the Consultant to consider in a consultation is how to progress with follow up.  There would be several options to consider:

  1. Review patient in several months for symptomatic review.  If despite 3 months of conservative management the symptoms were not responding, surgery could be considered at this time.
  2. Review patient in a longer time frame e.g. 6 months or a year.  This may avoid a further consultation in which the patient requires monitoring only and not progression to arthroplasty.  A safety net allowing the patient to contact the secretary sooner if symptoms worsen is an option.
  3. Discharge the patient from the clinic to allow the GP to continue to manage the condition until such time as symptoms worsen/progress, interfere with daily activities or no longer respond to conservative measures.

The last 2 options utilise secondary care resources more efficiently.  During my rotation I was consulting patients who travelled over an hour to get to the hospital; this may have involved a ferry journey from places such as Rothesay and Dunoon.  This costs the patient, or in some cases the health board, unnecessary expenses if the consultation would have been better suited in primary care.

NICE have also stated that patient referrals for consideration for arthroplasty should not be discriminated against due to factors such as age, smoking, obesity or other co-morbidities (5). This presents several interesting points.  One patient in the study did not go on to be listed for surgical intervention due to a body mass index (BMI) of 50.  The patient was deemed unsafe with a higher than normal risk of complications from surgery as a consequence of her BMI.  The patient was advised to consult her general practioner to pursue weight reduction programs coupled with dietary advice with subsequent re-referral after weight loss was achieved.  This made me think, should this have already been carried out before a referral to secondary care?  Whose responsibility is it to assess patient’s health and fitness for anaesthesia; primary or secondary care?  The patient felt disheartened and let down by the hospital.  She had waited 2-3 months for her appointment and was hopeful that seeing a specialist meant someone would offer her an operation.  After the consultation the patient stated,

‘You are doing a disservice to people who are overweight, I can’t walk because of the pain in my knees which means I can’t exercise so how can I lose weight?’ 

She was upset and angry at the consultant as she felt discriminated against.  I wondered if the patient had had the issue of her weight addressed by a doctor who she had an already established rapport with and therefore a level of trust, she may have been more understanding when the issue was discussed.  This would more likely be her GP than a secondary care consultant.  The patient did not want to discuss the risks associated with surgery and the reason behind the Consultants decision, the rapport here was lost.

Furthermore, if this patient had received adequate education and advice, as recommended by NICE, regarding weight reduction and the effects this could have she may not even require an operation in the future.  This also ties into one of our ethical principles – non malificience.   Sometimes patients may forget that an operation is not a simple undertaking and there are serious risks associated with it such as deep vein thrombosis, pulmonary embolus, myocardial infarction or even death (6).  As doctors we needed to do the best for our patient but also to remember to do no harm (7).

The second patient who was deemed not a candidate for surgery was elderly, living in a nursing home with severe dementia and not compliant with walking aids or medications.  Should this referral have occurred?  Or is this assessment and subsequent decision of the operating surgeon to deem who is appropriate, or not, for surgical intervention?  Perhaps there are other treatment options an orthopaedic surgeon could offer as an alternative?  They are specialists in the musculoskeletal medicine whereas a GP is a generalist who can only advise and manage a condition within the boundaries of his knowledge.

Three (14%) of the patients with OA did not wish to pursue surgical intervention.  Should these patients have been referred to secondary care or should their ideas, concerns and expectations be discussed in primary care to avoid unnecessary appointments?  Not every patient with osteoarthritis requires an operation if the natural history is fully understood.  Osteoarthritis does not inevitably get worse and in some patients their symptoms may improve in time (8).  As mentioned, exercise and weight loss may further help this (9). Arthritis Research UK state that a quarter of patients with osteoarthritis of the hip go on to have a hip replacement in 4 years from initial presentation to their GP (10).  Therefore, three quarters manage with conservative treatments alone. 

The audit also addressed whether investigations were carried out before referral.  Osteoarthritis is a clinical diagnosis but current guidelines advise that x-rays should be carried out before consideration for referral for arthroplasty is made (5).  This is twofold – to correlate symptoms with stage of disease but also to rule out other diagnoses (11).  Fifteen (71%) of the patients with OA already had x-rays carried out meeting this recommendation.  The six patients who did not have prior imaging experienced a longer wait in the outpatient department.  They would be seen by the Consultant, attend radiology for x-rays, return to wait for the images to be reviewed before a final discussion with the Consultant.  On a busy day, patients waited up to 3 hours to leave the clinic.  This caused frustration on both the patient and doctors part.  If appropriate imaging was carried out before referral this would make better use of secondary care time and resources as well as improving the patients hospital experience. 

So what do we need to do now, how can we improve things?  NICE emphasise the need for core conservative management options including education and advice for patients.  This is also needed for the healthcare professional.  As previously stated, it is possible to complete your GP training having not spent time in the orthopaedic department.  Furthermore a study carried out in Ireland in 2008 (12) assessed musculoskeletal knowledge of trainees including general practice trainees and qualified general practioners.  Forty (71%) GPs and 74 (71.8%) GP trainees were unable to achieve the pass score of 70.  Further review of their data showed that the pass rate was better for general practioners who had completed a musculoskeletal medicine rotation in the post graduate training (47.8% vs 18.1%, p <0.01) (12).  We need to be able to identify the gaps in our knowledge and ensure we address these to strive to become a competent GP (13). Some areas have introduced and piloted GP fellowship schemes as well as GP with specialist interests in musculoskeletal medicine (14).  But what about GP trainees?  Who should ensure we have the appropriate knowledge of common conditions in specialist areas such as this?  The individual as a self directed learner or our trainers as our educational supervisors ensuring we are competent to progress to certification as a GP?  A suggestion would be that those who do not have training in a set specialty identify themselves and arrange a session per week or month in a specialist clinic with a Consultant.  This was a method that qualified GPs felt would be helpful in their own learning also (1).  During meetings with our educational advisor we should be identifying areas of weakness and using private study sessions to address these. 

Effective management of osteoarthritis is achieved with a multi-disciplinary approach incorporating the GP, physiotherapist, orthotist, dietician along with orthopaedic surgeons and pain specialists (11).  Roddy et al (15) have piloted a scheme attempting to develop a ‘one stop clinic’ for musculoskeletal problems - a designated service for rapid assessment and treatment with access to radiology, medical devices and treatments including joint injections.  This would be advantageous in managing the condition as all members of the team are needed to provide optimal holistic care addressing the patients’ psychological, social and physical needs.  They found that the majority of patients in their cohort were managed in their clinic with only 10% of participants receiving onward referral to a specialist (either orthopaedics or rheumatology) (15).  This type of clinic may improve a patient’s experience of health care provision.

In line with NICE recommendations many geographical areas have produced referral pathways (16) as an aid for management of osteoarthritis but perhaps due to the clinician’s lack of awareness or knowledge of a condition, or pressures by the patient to see a specialist, these are not always followed.  Ensuring everyone has access to such pathways and are aware of how to utilise the referral pathway we could reduce unnecessary referrals.  The development of the referral letter itself with a box for different sections regarding the patient’s condition and management to date to be completed before referral is accepted is another possibility to improve the quality of referrals.  This may also allow vetting to be carried out and in place of review by the Consultant Orthopaedic surgeon, an extended scope physiotherapist assessment may be more appropriate.  (17)  Upgrading, or downgrading, of urgency of referrals could also be carried out more safely if a full history is presented.

In conclusion, osteoarthritis is a common condition that is faced by all health care practitioners in both primary and secondary care.  We need to ensure we provide a level of care outlined by NICE utilising the core conservative management options for patients and refer to secondary care when symptoms are interfering with their life or not responding to conservative management.  Improving both our own knowledge and providing education and advice for the patient to allow them to self manage osteoarthritis in the community.  We, as GP trainees, need to take some responsibility in our learning to identify any gaps in our knowledge and take the necessary steps in partnership with our supervisors to address these.  We need to continue this attitude into our GP career.  Appreciation of the importance of a good referral letter is crucial for transfer of patient information in an effective and resourceful way to improve both the patients experience and the relationship between primary and secondary care.

So to address the title question – if I was a patient who should I see with regards to my hip or my knee?  The answer is not as easy as I had first thought!  GPs are an ideal starting point acting as the gatekeeper to secondary care incorporating the NICE guidance and referral pathways, coupled with their clinical judgement to decide on the appropriate course of treatment for a patient.  But they cannot work alone.  We all need to work together within the multi-disciplinary team with other health professionals such as orthopaedic surgeons, physiotherapist, orthotist, chronic pain teams and dietician to manage osteoarthritis.  In a ‘perfect world’ we would have one stop joint problem clinics with specialist GPs, physiotherapists, pain team and orthotic input to provide a holistic model of care for our patients. 



  1. Roberts C, Adebajo O, Long S.  Improving the quality of care of musculoskeletal conditions in primary care.  Rheumatology 2002;41:503-508
  2. Booth A, Wise DI.  General Practice training in musculoskeletal disorders.  British Journal of General Practice.  1990;40:390
  3. Newton J, Eccles M and Hutchinson A.  Communication between general practioners and Consultants: what should their letters contain?  BMJ 1992;304:821-4
  4. Harvard Team.  Special Report No. 5, Honk Kong Patient Studies Summary Report,  improving Hong Kongs Health care System: Why and For Whom?, 1999
  5. National Institute for Health and Clinical Excellence. Osteoarthritis: national clinical guideline for care and management in adults . London: NICE, 2008.  (accessed 7/12/13)
  6. Gidwani S, Fairbank A.  The orthopaedic approach to managing osteoarthritis of the knee.  BMJ 2004;329:1220-4
  7. General Medical Council.  (2013)  Good Medical Practice. GMC (accessed 28/04/13)
  8. Felson, D.T. Developments in the clinical understanding of osteoarthritis. Arthritis Res Ther, 2009. 11(1): p. 203
  9. Peters, T.J et al.  Factors associated with change in pain and disability over time: a community-based prospective observational study of hip and knee osteoarthritis. British Journal of General Practice, 2005. 55(512): p. 205-11
  10. Arthritis Research UK.  Osteoarthritis: a modern approach to diagnosis and management.  Reports on the Rheumatic Disease, Series 6, Autumn 2011, Hands on No 10
  11. Hunter DJ, Felson DT.  Osteoarthritis.  BMJ 2006;332:639-42
  12. Queally JM et al.  Deficiencies in the education of musculoskeletal medicine in Ireland.  Irish Journal of Medical Science.  June 2008;117:99-105
  13. Akesson K, Karsten E, Woolf AD.  Improved education in musculoskeletal conditions is necessary for all doctors.  Bulletin of the World Health Organization 2003;81 (9)
  14. Duckett S, Casserly H.  Orthopaedic GP fellowship:o   does it work?  Ann R Coll Surg Engl 2003;85: 195-196
  15. Roddy E, Zwierska I, Jordan K et al.  Musculoskeletal clinical assessment and treatment services at the primaru-secondary care interface: an observational study.  British Journal of General Practice 2013;63:87-88
  16. Scottish Governmebt Task and Finish Group.  Musculoskeletal – Knee Pain Referral and Management Pathway.  2011. (Accessed 17/5/13)
  17. Daker-White G, Carr AJ, Harvey et al.  A randomised controlled trial.  Shifting boundaries of doctors and physiotherapists in orthopaedic outpatient departments.  J Epidemiol Community Health 1999;53:642-650



Appendix 1 – Referral Letter Proforma



Name of GP –

Date of referral –

Date of outpatient visit –

Urgency- routine / Urgent

Presenting complaint – Adequate details of location of the problem

Yes    /    No    ____________________________

Duration of symptoms


1. Not given

2. < 6 weeks

3. 6 weeks to 6 months

4. 1 year – 5 years

5. > 5 years


Provisional GP Diagnosis

Yes    /    No  ____________________________

Is the patient known to orthopaedics for the same complaint?


Yes and documented in the letter

Yes and not documented in the letter


If yes to above, is the previous Consultant documented in the letter?

Yes    /    No

Are adequate examination findings provided in letter

Yes    /    No

What non operative treatments have been attempted prior to referral

1. No information provided

2. Analgesia

3. Walking aids

4. Physiotherapy

5. Injections

6. Not req. before referral

What investigations have been done prior to referral?


1. Blood tests

2. X-rays

3. MRI

4. Ultrasound

5. No information provided

Orthopaedic Department Diagnosis


Outcome of Orthopaedic Consultation

1. Physiotherapy                                                                 

2. Analgesia

3. Surgery

4. Referral to other consultant

5. Patient declined intervention

6. Imaging

7. DNA

Opinion of Orthopaedic Surgeon regarding appropriateness of referral to Specialist


Yes it was appropriate


No it was not appropriate