You can’t have your op till you’ve lost weight  

It’s back

With a vengeance

Like all good zombies this one crops up every few years when money’s tight. There’s unending beleif that if only we implemented a BMI threshold for elective surgery we would save big.

Personally I think this is wrong, generally crackers, difficult to substantiate with evidence, against the NHS constitution and a route to trouble.

Last time I had to write something on this was some years ago. I wheel it out again every few years. Sometimes I think I might need to update it, but as of yet I can’t summon the will. Here is the 2011 paper. Smoking will follow next.

 

What is the evidence on BMI affecting short, medium and long term outcomes in JRS? 

Summary of Key points from Clinical Evidence on this question

· Difficult to draw substantive conclusions

· The quality of research reviewed was mixed

· Some reasonably large studies – 100 – 600 patients. Presumably based on audit of patient cohort

· Little of the evidence reviewed cites outcomes in a full epidemiological context (eg person years or some such measure that includes a time element). Outcomes must be viewed in the context of the whole population, rather than patient by patient. Absolute measures of outcome are important.

· Where there are outcome measures quantified; it is frequent that these are reported as relative measures (relative risk, hazard ratio etc). Without knowing the baseline rate, it is thus hard to make firm conclusions as to the absolute risk of poor outcomes associated with poor outcomes in obese populations – ie what happens in whole populations.

· Although there is a trend towards poorer outcomes, it is hard to be definitive about this. There is much conflicting research.

· Much of the published research does draw a distinction between obesity and morbid obesity – the findings on MO are a little sharper.

· From the basis of the available evidence it is difficult to draw a different conclusion to NICE – that “BMI should not be a barrier to referral for joint replacement surgery”

· NICE CG has explicitly precluded obesity as a barrier to an op, in this respect we would be explicitly going against NICE

 

Contents

1. Context and background

2. Clinical evidence

a) NICE

b) Primary research

c) Unanswered questions

3. Views of others

a) Other PCTs

b) Orthopods views

4. Ethical considerations

5. other issues

a) GMC guidelines on Good Medical Practice

b) Reputation and comms

6. options
1 Context and background

Ongoing discussion re the pros and cons of putting in a threshold for MO patients.

Mixed clinical evidence

Mixed views from different clinicians

Obesity increases risks and complications – anaesthetics and surgical / QoL. 

Reputationally, clinically and ethically difficult.

BMI 40 wont have much of an impact; do we need to discuss a BMI threshold lower than 40?

 

Weight reduction prior to surgery might be particularly good for knee problems. It helps in delaying the requirement for surgery. In moderate cases, some patients may never require to have knee replacements. Knee replacements may be avoided by weight loss / quad strengthening.

NICE CG has explicitly precluded obesity as a barrier to an op, in this respect we would be explicitly going against NICE

 

There are outstanding uncertainties re the clinical and ethics of restricting on BMI

No mention of BMI in the regional threshold work

If we set a BMI threshold, what would it be 35 or 40, or 37

 
2 Clinical Evidence

a) NICE view

NICE CG59 states categorically that BMI shouldn’t be used for clinical decision making

R35 Patient-specific factors (including age, gender, smoking, obesity and comorbidities) should not be barriers to referral for joint replacement therapy.

There is a difference between clinical decision (at patient level) and a commissioning one, PCTs are not bound to follow or implement CGs. It might be difficult to explain that to the local media in the current climate.

 

b) Primary Research

Reasonably comprehensive lit search was undertaken to consider whether there is published clinical evidence of differences in outcomes between normal weight, obese and morbidly obese.

Articles published in the last 10y were considered

For brevity, only abstracts reviewed and an overview of this data is provided here, rather than a full synthesis and summary of evidence.

 

General issues with the evidence reviewed

· Obesity IS a risk factor for OA and later TKA / THA.

· Significant uncertainty in the published evidence as to whether outcomes are worse among obese patients.

· Although there is, perhaps, a tendency amongst clinicians that outcomes are less good in obese and MO; it is fair to say that there is still significant uncertainty in the literature. There is no clear consensus in the literature.

· Difficult to draw substantive conclusions

· The quality of research reviewed was mixed

· Some reasonably large studies – 100 – 600 patients. Presumably based on audit of patient cohort, some with (and some without) matcing controls

· There is a greater reporting / research base in knees compared to hips

· Little of the evidence reviewed cites outcomes in a full epidemiological context (eg person years or some such measure that includes a time element). Outcomes must be viewed in the context of the whole population, rather than patient by patient. Absolute measures of outcome are important.

· Some of the research is cited as with, some without, comparator groups.

· Where there are outcome measures quantified; it is frequent that these are reported as relative measures (relative risk, hazard ratio etc). Without knowing the baseline rate, it is thus hard to make firm conclusions as to the absolute risk of poor outcomes associated with poor outcomes in obese populations – ie what happens in whole populations.

· Few if any of the studies had a medium or long term economic component – we do not know whether there is a differential in incremental cost effectiveness of TKA / THA between obese / non obese. Where economic variables were collected it was short term – length of stay etc. Perhaps a tendency to conclude that although obese and MO patients can and do benefit from JRS, the benefit is slightly less than for normal weight, and the short and long term cost is less (ie less cost effective).

· Many of the studies may not have corrected for obvious confounding variables. For example outcomes may easily be confounded by specific type of procedure (eg cement v cementless); operative technique / operator error; age; gender etc.

· It is thus hard to definitively quantify whether outcomes of TKA / THA are poorer in obese patients, as such clinicians still need to weight up the risks and benefits of each case.

· Although there is a trend towards poorer outcomes, it is hard to be definitive about this. There is much conflicting research.

· Much of the published research does draw a distinction between obesity and morbid obesity – the findings on MO are a little sharper.

· From the basis of the available evidence it is difficult to draw a different conclusion to NICE – that “BMI should not be a barrier to referral for joint replacement surgery”

 

Outcomes considered and my take on General findings from the literature

Pre op and operative
Anaesthetic risks / technical challenge / choice of prosthetic / operative time / LOS

More technically challenging procedure

No strong data confirming longer LOS in obese, some evidence re no difference.

Post op

wound infection, DVT, complications of surgery,

BMI has weak correlation to early outcomes (may be that operator factors have a stronger correlation to early outcomes).

Higher complication rate (esp DVT and wound infection); higher health care utilisation, poorer improvements in knee or hip score and pain scoring.

Early

satisfaction, clinical scores, pain scores, rehab, mobility

Some limited research compared functional status – walking speed, flexion, mobility….. limited eviodence re differences between obese / non obese in rehabilitation outcomes

Reasonable evidence that obesity has a negative effect on clinical outcomes (pain scores, other clinical scoring) – esp in short term post op

Limited to no conclusive evidence re difference in pain scoring up to 5y or HHS up to 5y. Where evidence was found, there was a slightly lower improvement in reported scores.

Up to 5y patient satisfaction is slightly lower in obese

Late

revision, prosthetic survival, clinical scores, pain scores, rehab, implant dislocation, mobility, functional gains

Maybe slight tendency towards greater rate of implant dislocation.

Maybe slight tendency towards shorter survival of prosthetic (and thus higher revision rate). Whilst in the context of the individual patient this is not important, but aggregated across a population of patients over a long period of time, this is an important factor. It is hard, however to draw a definitive conclusion on this – the evidence is contradictory in parts.

Recovery and long term rehabilitation will occur in obese patients but the absolute gain may be less, slower and at higher expense.

Mixed findings re weight loss following surgery, revision rate, prosthetic survival, ongoing pain control meds

No evidence re mobility

 

c) Unanswered questions include

· What are the long-term results of a total hip arthroplasty in morbidly obese patients?

· Given the rather gradual increase in symptoms with obesity, is it worth asking obese patients to lose weight prior to

· total hip replacement?

· Should anticoagulation regimens be different in obese women undergoing THR?

 

 

 
3 Views of others

a) Other PCTs

We should not simply unthinkingly follow what other PCTs have done.

Some PCTs have set explicit BMI thresholds. Relatively few have cited any evidence that underpins their deliberations. In those PCTs that have, it is impossible to establish how comprehensive an evidence review was undertaken – it is readily possible to selectively cite evidence in order to justify a pre ordained position.

Some of the PCTs have stopped short of an explicit threshold but have made clear that ALL efforts to reduce BMI to <30 in the x months prior to referral should have been trailed and failed. Obviously this implies documented compliance with evidence based weight management interventions; rather than a simple brief intervention, or worse a ‘casual chat’.

Some PCTs have stated that a BMI of (x; 35 – 37)or over does not absolutely preclude referral but all reasonable attempts should be made to reduce weight to levels suitable for surgery as higher BMI carries increased risks of bleeding and post-operative thromboembolism.

 

 

 

 

b) Secondary care clinicians views

Ortho surgeons

Our orthos are of view that you can come up with enough evidence to suit any viewpoint the commissioner might wish to take – prosthetic survival, ST and LT complications, restoration of functionality.

 

By setting a threshold we would need to consider the global loss of QoL (in the patients that would be denied JRS) and spend in other areas

 

Anaesthetistic risk

This is uncertain. Views of anaesthetists on operating with high BMI have not been tested systematically.
4 Ethical considerations

This is likely to be a deeply ethical issue, with much contention. There are two sides to the debate that merit equal consideration
Ethical issues in favour of using BMI to limit access to JRS

• Failiure to loose weight = clinical risk in surgery – patients putting themselves at risk

• When all other clinical features are identical, costs are increased and outcomes are worse in obese than in non obese. (NB the clinical evidence does not bear this out in this case, though it does in others – eg smoking / surgical risk)

• Utilitarian approach – in system with limited £ – > outcomes if less obese are prioritised for JRS. It is less cost effective to operate in obese (thus net social loss of health)

• Could be viewed as a sensible clinical judgement to use BMI as a discerning factor (esp BMI>40). Complications, healing, infections, long term outcomes, revision rate. Compromises outcome and increased cost of care.

• Opportunity cost of treating obese with JRS (with their > complication rate and LOS and poorer / costlier outcomes) = less opportunity to treat other patients – thus net loss of health. Again evidence is mixed in this specific example.

 

Ethical issues against using BMI to limit access to JRS

• undermine the doctor-patient relationship – may significantly harm future relationship and patient may not disclose other info.

• Doctors routinely treat patients who are despised by the society in which they live—enemy troops, terrorists, murderers – what is so bad about obese patients. Obesity is very common.

• Simple discrimination or victimisation of obese. Who may not be able to loose weight (whose responsibility is obesity – individual or social).

• How to classify obese and risk of poor outcomes (uncertain what the “correct” threshold for BMI would be.

• Most obese have good outcomes (ST and LT) – although there is a trend towards poorer outcomes, when the risks are quantified in relative AND absolute terms, the risk for most obese patients might be low. Only a few (but a higher proportion than non obese) would have poorer outcomes. Again, in a clinical context this is relatively important.

• Thin end of wedge. Obesity next, then drinking, then where. May put serious limitations on the notion of liberty. Should we treat a person with a sports injury – after all it is self inflicted

• Denying care to a group we have previously provided care to – because money is tight.

• Significant equity concerns – disadvantaged are more likely to be obese – therefore agreeing a BMI threshold does hold significant inequity potential.

• Most clinicians would see the primary duty of the doctor to be their patient(s) rather than populations? Doctors owe a primary duty to the patient rather than the population (BUT how long can this untrammelled advocacy for the individual survive)?

 

May be a group of MO patient with no hope of weight loss, severe symptoms (night/rest) and don’t meet criteria for bariatric surgery – ?the geriatric population. One could make a case that ethically unacceptable to not offer them surgery so long as they consent to the significant increased risks. This is a very small group as there are very few old morbidly obese patients for obvious reasons!

 

 

 

 

 

 

5 Other issues

a) General Medical Council guidelines on good medical practice:

 

• (para 9). You must give priority to the investigation and treatment of patients on the basis of clinical need, when such decisions are within your power. If inadequate resources, policies, or systems prevent you from doing this, and patient safety is or may be seriously compromised, you must follow the guidance in paragraph 6

• Para 6. Draw the matter to the attention of [my] employing or contracting body.” This is what I have done and will continue to do, in more discussions than I can remember with primary care trusts, commissioners, and within the trust, and in the drafting of one of many (not yet accepted) business plans

 

 

 

b) Reputation and comms

this issue will inevitably end up on local media. Repulatational risk management is an important consideration    

 

Any move towards setting an explixit BMI threshold must pass an “Accountability for Reasonableness” test

This framework identifies four conditions for seeking fairness in health priorities:

1. Relevance – the rationales for priority setting decisions must be based on reasons (evidence and values) that stakeholders can agree are relevant. Procedurally, having a wide range of stakeholders participate in deliberation ensures that the full range of relevant reasons will be considered.

2. Publicity – priority setting decisions and their rationales must be publicly accessible, not just on demand, but through various forms of active communication outreach.

3. Revisability – there must be processes for revising decisions and policies in response to new evidence, individual considerations, and public reactions.

4. Enforcement – local systems and leaders must ensure that the above three conditions are met.

 

 
6 Options

NICE CG has explicitly precluded obesity as a barrier to an op, in this respect we would be explicitly going against NICE

There are outstanding uncertainties re the clinical and ethics of restricting on BMI

No mention of BMI in the regional threshold work

If we set a BMI threshold, what would it be 35 or 40, or 37

Do we stick at 40, or “go low”. If we do go low then we need to do so understanding the clinical, ethical, equity and reputational impacts for consideration in detail by our clinical executive. BMI at <36 would be rationing.

 

If we were to implement a BMI threshold

Consider something along the lines of:

 

“a BMI of (x; 35 – 37)or over does not absolutely preclude referral but all reasonable (and documented) attempts should be made to reduce weight to levels suitable for surgery as higher BMI carries increased risks of bleeding and post-operative thromboembolism in addition to a concern about poorer long term outcomes”.

 

AND

 

“BMI > x (?37) then either auto referral to physio and weight mx or no referral to ortho unless recent, documented referral to weight management and failure to loose weight….over xxxxxxx months”

 

Exceptions to such a policy might include:

· patients whose pain is so severe and/or mobility so compromised that they are in immediate danger of losing their independence and that joint replacement would relieve this threat;

· patients in whom the destruction of their joint is of such severity that delaying surgical correction would increase the technical difficulty of the procedure.

 I do think this is difficult to justify

 

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s