Rapid Responses to:
|
|
Rapid Responses: Rapid Responses published:
-
![[Read Rapid Response]](/icons/misc/sectionDOWN.gif)
BMJ readers should be even more cautious than usual in interpreting the results of the Evercare evaluation
- Peter Yuen (15 November 2006)
-
Redesigning primary care to reduce unplanned hospital admissions
- Nancy Kay Williams, Gravelle, et al (15 November 2006)
-
Community Matrons can make a difference
- Christopher S Jenner, Nina L Barnett, Specialist Pharmacist for Older People (21 November 2006)
-
Fundamental healthcare reforms should be driven by evidence for beneficial outcomes rather than reduction in unscheduled care costs.
- Nicholas I Church (21 November 2006)
-
BMJ readers should be even more cautious than usual in interpreting the opinions of UHE employees
- Andrew Clegg (23 November 2006)
-
How about a trial?
- Peter A West (24 November 2006)
-
Case Management. Where is the government's committment to evidence? How can United Healthcare remain impartial?
- David Oliver (25 November 2006)
-
Measured responses to evaluations of case management
- Anthony P Roberts, Lucy Harding (30 November 2006)
-
Community Matrons do make a difference
- Martin J Howard (1 December 2006)
-
Why evidence does matter
- David Oliver (2 December 2006)
-
'Avin a larf..
- david leopold (2 December 2006)
-
Let's develop our own model of case management
- Nicole L Klynman, Ugo Okoli, Director of Public Health, Enfield Primary Care Trust (7 December 2006)
-
Re: Glass half full?
- Louise Gibson, community matron students, University of Chester (7 December 2006)
-
Re: Re: Glass half full?
- Andrew Clegg (10 December 2006)
-
The baby, the bathwater and the missing RCT
- Rowan H Harwood (11 December 2006)
-
Five questions which need answers
- david oliver (12 December 2006)
-
Case management can reduce the costs of unplanned admissions
- Wendy A M Walker (5 January 2007)
-
what is a saving?
- john sharvill (7 January 2007)
-
Strategies to improve the evercare programme
- S Kapoor M.D. (9 January 2007)
-
Authors' response
- Martin Roland, Hugh Gravelle, Mark Dusheiko, Rod Sheaff, Penny Sargent, Ruth Boaden, Susan Pickard, Stuart Parker. (31 January 2007)
-
Implementing Case Management – understanding principles and learnings from complex adaptive systems.
- Carmel M Martin, Joachim P Sturmberg (7 February 2007)
|
|
|||
|
Peter Yuen, Director of Analytics UnitedHealth Europe, 15 Greycoat Place, London Sw1P 2SB
Send response to journal:
|
Gravelle and others spell out some of the many methodological weaknesses in their evaluation of the Evercare pilots, but I thought that it might be helpful to readers to list all that they mention and some others. It’s a formidable list. 1. As your editorial says, a randomised controlled trial is the best way to evaluate the effectiveness of an evaluation, and this study is a very long way from being such a trial. 2. The “control” group were not selected in the same way as the Evercare group—and so this study is not comparing like with like. 3. The study has severely limited power—so a sizeable reduction in unplanned admissions could have been missed. 4. The authors couldn’t identify individual patients who were in the Evercare programme. Instead, they had to study practices that were part of the programme, but some of those practices may have had very few patients in the programme. In other words, the practice may have had a very “low dose” of the intervention. 5. Furthermore, Evercare practices were different from “control” practices. The Evercare practices had bigger lists, were more deprived, and had more patients at higher levels of risk. In particular, at the beginning of the study they had higher rates of unplanned admissions. Again like is not being compared with like. 6. The average length of time that patients were in the programme was eight months, which is almost certainly too short a time to see an effect. Evidence from the US suggests that a programme needs to be running for 18 months to see full benefits. 7. As the authors acknowledge, when patients were first entered into the programme nurses may have identified problems that demanded hospital admission. 8. The “control” practices could have included practices that were using other forms of case management, causing contamination and complicating interpretation of the results. 9. The paper reports a higher mortality among the intervention group. But as their data source (Hospital Episode Statistics) do not count all the deaths outside hospitals, as the authors point out, is mortality a relevant outcome measure? Plus there was never any expectation that Evercare would reduce mortality. I wonder if this is not a case of publication bias. The problem whereby negative studies are not published because they are deemed boring is well known. Here it may be the other way round. It’s hard to believe that the BMJ would have published such a weak study if the result was positive. The fact that it was negative makes the study more newsworthy—and perhaps more tempting to the editors with their journalistic instincts. Peter Yuen Competing interests: Peter Yuen is an employee of UnitedHealth Europe, the company that conducted the Evercare pilots. |
|||
|
|
|||
|
Nancy Kay Williams, Clinical Director UnitedHealth Europe SW1, Gravelle, et al
Send response to journal:
|
Redesigning primary care to reduce unplanned hospital admissions Gravelle and others conclude in their study that radical system design is needed to reduce unplanned hospital admissions, (1) and UnitedHealth Europe UHE agrees. UHE has worked with many primary care trusts to help them implement their programmes of improving the care of patients with long term conditions, and we have identified some 16 features that are needed to make those programmes successful in reducing unplanned hospital. The features are shown in the box, but I want to enlarge on three. Firstly, a reliable information tool is needed to identify the right patients to be case managed. The interim evaluation by showed that “two unplanned admissions in patients over 65” is a poor predictor of unplanned admission. (2) It was all that could be used in 2002 when the Evercare pilots began, but the RISC tool that we have developed and installed in many PCTs is a much better predictor. Secondly, there must be a system for covering patients out of hours as that’s when most unplanned admissions occur. Again this wasn’t possible with the original pilot. Cover might be achieved through providing the community matron service round the clock or by making careful links with out of hours services—in the way that some palliative care services do. Thirdly, it’s necessary for commissioners to ensure that alternatives such as step up intermediate care beds are made available. These were not available in the Evercare pilots. Without such beds frail, elderly patients who cannot be left in their homes but do not need the full services of an acute hospital will have to be admitted. PCTs must find ways to reduce unplanned hospital admissions, and we are confident that one way to do so is to radically redesign the system—with well trained community matrons being one part of the redesign. Competing interest: NW is an employee of UnitedHealth Europe, the company that conducted the Evercare pilots. Box: 16 features that a programme of improving the care of patients with long term conditions needs in order to succeed in reducing unplanned hospital admissions 1. A clear vision of what the programme is for, how it will work, and
what it must achieve (better care, reduced admissions, or both?)
Competing interests: Nancy K WIlliams is an employee of UnitedHealth Europe, the company that conducted the Evercare pilots. |
|||
|
|
|||
|
Christopher S Jenner, General Practitioner, Older People Champion, Harrow Elliott Hall Medical Centre 165-167 Uxbridge Road Hatch End, Pinner HA5 4EA, Nina L Barnett, Specialist Pharmacist for Older People
Send response to journal:
|
We read with interest the paper by Gravelle et al and the accompanying editorial by Black regarding the findings of the Evercare Pilot (1,2). The paper by Gravelle et al suggests that community matrons working within the Evercare Pilot do not reduce emergency admissions to hospital. These findings have far reaching consequences for the NHS and for patients and we would like to contribute to the reflection on this subject with observations from our own successful, locally run, community matron pilot. Gravelle et al suggest that a better method for identifying high risk groups is warranted. We observed this early on in our pilot and adapted our methods for identifying Very high Intensity users (VHIUs). This included a combination of PARR (Patients At Risk of Readmission) data (version 1), (3) locally adapted criteria for VHIU identification (4) ,information from members of the individual practices and from community practitioners. Gravelle et al’s study included any practice involved in the pilot with patients over 65 years with two or more emergency admissions in the previous 13 months. We adopted an alternative practice based approach, identifying initially all VHIUs of 85 years and over in pilot practices as we found this age group to have the highest number of VHIUs. The practice based approach was key to integrating the community matron into the primary health care team and ensuring that all General Practice team members were engaged in the process. Central to this approach has been the use of the GP IT systems as the primary source of record keeping facilitating both communication and clinical governance. Matrons also followed their patients across health and social care boundaries ensuring communication and continuity of care planning with formal and informal carers. We noted that stakeholder engagement at the start of the process was important to avoid the potential for antagonism with other health care professionals through the process of change. Black suggests that using UnitedHealth Group recruited nurses and waiting longer for results to become apparent may have altered the outcome of the Evercare pilot. Our work reflects this view and has shown that the lag time to benefit is a minimum of 1 year, something that we did not predict at the start of the pilot. Our nurses required substantial education and training for the advanced practitioner role (6,7) and support from multi-agency mentorship throughout the pilot. Vulnerable older people in the community require support from a wide range of health and social care professionals in primary and secondary care. The role of a nurse with advanced clinical skills is central in facilitating coordinated care for vulnerable older people and providing clinical supervision for other nurses undertaking this care. Service redesign may be required in areas where a gap in provision is identified as was the case in our locality. We believe that a flexible locally adapted initiative focussing on case management of this client group can provide benefits for the patient and the NHS. Christopher S Jenner FRCP FRCGP (Chair) and Nina L Barnett MSc MRPharmS (medicines lead) on behalf of the community matrons steering group, Harrow Primary Care Trust. References 1. Hugh Gravelle, Mark Dusheiko, Rod Sheaff, Penny Sargent, Ruth Boaden, Susan Pickard, Stuart Parker, and Martin Roland Impact of case management (Evercare) on frail elderly patients: controlled before and after analysis of quantitative outcome data BMJ, Nov 2006; doi:10.1136/bmj.39020.413310.55 2. David A Black. Case management for elderly people in the community BMJ, Nov 2006; doi:10.1136/bmj.39027.550324.47 3. Patients at Risk of Re-hospitalisation (PARR) case finding tool. Kings fund. http://www.kingsfund.org.uk/health_topics/patients_at_risk/index.html 4. Castlefields Heath Centre. Chronic Disease Management. 2004 http://www.natpact.nhs.uk/eventmanager/uploads/castlefields.ppt 5. Department of Health. April 2006. Caring for people with long term conditions:an education framework for community matrons and case managers. http://www.dh.gov.uk/assetRoot/04/13/40/12/04134012.pdf 6. Department of Health. Feb 2005. Supporting people with longterm conditions:liberating the talents of nurses who care for people with longterm conditions. 7. Department of health. Aug 2005. Case Management Competencies Framework. Competing interests: None declared |
|||
|
|
|||
|
Nicholas I Church, Consultant Physician and Gastroenterologist Queen Margaret Hospital, dunfermline, Fife. KY12 0SU
Send response to journal:
|
Editor The results of the paper by Gravelle et al (1) have already been misleadingly reported in the medical press. The Evercare pilot scheme, a community based, nurse-led case management programme designed to reduce unplanned hospital admissions in frail elderly people, is stated to have increased emergency admissions by 2.5% and mortality by 5.5%, while reducing emergency bed days by 4.9%. (2) Further “headline grabbing” statements claiming that these results undermine heath care reforms aiming to reduce hospital admissions are likely to follow. In fact, the differences between the intervention and control groups in the Gravelle paper are not statistically significant, the confidence intervals are wide, the groups were not matched and this was not a randomised trial. The methodological limitations of the paper are well described by Yuen in his rapid response (conflict of interest noted). (3) Despite the weaknesses of the Gravelle study, it serves to illustrate a wider point. Health care delivery has undergone fundamental changes in recent years, many of which are based on well meaning ideas rather than evidence of effect. It is assumed that avoiding emergency hospital admission is good for patients, but is this really the case? Are the outcomes of care in the community better? Black, in the accompanying editorial, identifies the lack of hard evidence in this area. (4) As health care providers, we are all under intense pressure to reduce unscheduled care costs by rapidly discharging patients from hospital, or by preventing them from coming in. However, before we advocate yet more reforms, we should take a critical look at whether we are actually improving the health of patients or simply moving the problems elsewhere. References: 1 Impact of case management (Evercare) on frail elderly patients: controlled before and after analysis of quantitative outcome data Hugh Gravelle, Mark Dusheiko, Rod Sheaff, Penny Sargent, Ruth Boaden, Susan Pickard, Stuart Parker, Martin Roland BMJ, doi:10.1136/bmj.39020.413310.55 (published 15 November 2006) 2 Matrons do not cut hospital admissions. Rachel Liddle. GP newspaper; 2006 (17 Nov): 1 3 BMJ readers should be even more cautious than usual in interpreting the results of the Evercare evaluation Peter Yuen (15 November 2006) 4 Case management for elderly people in the community David A Black BMJ, doi:10.1136/bmj.39027.550324.47 (published 15 November 2006) Competing interests: None declared |
|||
|
|
|||
|
Andrew Clegg, SpR Geriatrics Yorkshire Deanery
Send response to journal:
|
I read with interest the reponse by Peter Yuen (1), director of analytics at UnitedHealth Europe, in reaction to the results presented by Gravelle et al (2), namely that there is no significant difference in rates of emergency admission, emergency bed occupancy and mortality between the control and intervention group. I feel that he may wish to congratulate the authors for correctly presenting the data in a neutral light, as opposed to showing a non-significant increase in rates of admission, bed occupancy and mortality in the intervention group. Mr Yuen criticises the statistical basis of the trial, citing the fact tht the trial is a 'very long way' from being a randomised controlled trial. This statement is let down by the fact that his own 'formidable list' of methodological weaknesses is largely a stream of hypotheses generated on the basis of his own speculation and conjecture. I would like to make two specific points regarding his list. Firstly the limited power of the study 'may' also have missed a sizeable increase in unplanned admissions in the intervention arm of the trial. Secondly, entry into the programme 'may' have led to identification of problems necessitating hospital admission. This would still equate to a tangible increase in hospital admissions. One may view the final paragraph with a degree of cynicism. Before criticising the BMJ for publishing a 'negative study', I feel that Mr Yuen should consider his reaction to this 'methodologically flawed trial' had it come to the conclusion that there was a statistically significant reduction in hospital admissions in the intervention arm. Instead of his string of criticisms I feel that UHE would have been singing the praises of the authors and (metaphorically speaking) ramming the results of the study down our throats. He may wish to review the arguably hypocritical statements that are laid out in this paragraph. Furthermore, I would like to draw the attention of Mr Yuen to the editorial by David Black (3). If he truly wishes to practice evidence based medicine he would do well to appraise the meta-analysis by Stuck et al (4) which provides the basis for the statement that 'comprehensive geriatric assessment as an inpatient, with ongoing control over medical recommendations, remains the proved intervention as it reduces mortality, reduces institutionalisation, and improves functional status'. Finally I would like to point out an omission by Miss Williams (5) with regard to her pompous, inane mission statement of '16 points'. I commend to you point 17. If a programme to improve the care of patients with long term conditions is to be truly successful it should also require 'Evidence that the programme actually works'. 1. Yuen P. BMJ readers should be even more cautious than usual in interpreting the results of the Evercare evaluation. BMJ Rapid Responses. Nov 15 2006. 2. Gravelle H et al. Impact of case management (Evercare) on frail elderly patients: controlled before and after analysis of quantitative outcome data. BMJ 2006; Nov 15. Online First. 0: bmj.39020.413310.55v1. 3. Black DA. Case management for elderly people in the community. BMJ 2006; Nov 15. Online First. 4. Stuck AE et al. Comprehensive geriatric assessment: a meta- analysis of controlled trials. Lancet 1993; 342: 1032-6. 5. Williams N. Redesigning primary care to reduce unplanned hospital admissions. BMJ Rapid Responses. Nov 15 2006. Competing interests: I am not employed by a company with a vested interest in a positive outcome to the study |
|||
|
|
|||
|
Peter A West, Senior Research Associate York Health Economics Consortium (YHEC) University of York
Send response to journal:
|
Gravelle et al.'s study of the Evercare initiative has provoked some critical comments from the company and from NHS staff with knowledge of the field. Could the matter not be solved by asking that companies such as Evercare undertake the necessary trials of interventions such as this, from which they anticipate making a profit. We require this standard from pharmaceutical companies where the cost per patient may be comparable. If the company mounted suitable trials, with the necessary safeguards for independence etc. then it would not need to criticise what the study authors accept was not a perfect trial. Rather, it could put forward its own high standard evidence of effectiveness (or not) in an English or UK NHS setting. Competing interests: YHEC is a contract research company owned and operated by the University of York. It provides research and consultancy for the NHS and the private sector. The author is currently on a long term secondment to a government agency. No conflict of interest has been identified. |
|||
|
|
|||
|
David Oliver, Senior Lecturer and Hon Consultant Geriatric Medicine Institute of Health Sciences, University of Reading, London Road, RG1 5AQ
Send response to journal:
|
Sir Louise Casey, the government’s “respect tsar” epitomised disrespect, when she said “ if number 10 says bloody evidence-based policy to me one more time, I’ll deck them…who needs evidence when you have politics?”. [1] When we consider several recent initiatives in healthcare, this seems a surprising statement. Many elements of “Your Health, Your care Your say” [2]and the NSF for long term conditions [3] are in fact poorly evidenced. This is not an esoteric matter as the cost and opportunity costs associated with implementing eye catching but poorly-evidenced service initiatives based on hunches or political imperatives will compromise other aspects of patient care. Simultaneously, there are many examples of well-evidenced interventions (e.g. for falls, osteoporosis, or stroke) not being delivered to the majority of patients likely to benefit. [4]. Few clinicians would argue against the necessity to focus on more joined up, proactive care of persons with multiple chronic conditions and the likely benefits for patient care or the wider health system, but the Evercare story is a lesson on how not to go about this. The Evercare pilot was based on the findings of one quasi- experimental study of case management in US nursing homes, albeit one with impressive results. [5] It was imported to the UK in ten pilot sites and applied to a very different group of patients (community dwelling) , in a very different health economy, with different nursing skills and information system. Even the author of the original paper expressed surprise that the intervention had been implemented in this way. [6] This cost the UK taxpayer over £4m – not to mention the cost of “backfill” for community nursing posts vacated by the new advanced practitioner nurses. If the Department of Health had any regard for evidence it would not have ignored two excellent systematic reviews [7.8] which showed no consistent evidence for the effectiveness of case management in preventing hospital attendances or admissions, reducing health costs or improving function. Nor would they have repeatedly touted the “Castlefields” model [9] (never published in a peer reviewed journal) as good evidence for chronic disease management. Most importantly, they would have commissioned a good RCT of the intervention in the UK instead of commissioning an evaluation almost as an afterthought. Turning to the criticisms from Yuen [10] and Williams [11] in this journal – both of whom are senior employees of United Healthcare and have rightly declared their fundamental conflict of interest they are quite right to point out certain methodological limitations in the study. However, these same limitations were more than apparent in United Healthcare’s own interim (and independent?) report on the Evercare project – [12] which with no peer review, no tests of statistical significance and an even shorter duration of intervention ran to several pages in praising the scheme’s apparent success and which for some reason was allowed to appear on the Department of Health’s own website. Nor to my knowledge, when United Healthcare were offered the contract by the DH for the pilots advise the government to wait and conduct the very RCT that Yuen now advocates should have been conducted and to build in more robust evaluation from the start. Rather they were happy to provide the consultancy and training packages to the participating Primary Care Trusts. This is doubly concerning in view of the vast sums spent by the current government on external consultancy. [13] I believe that the only parties to emerge with credit from this exercise are the research team who did their best to perform a quasi experimental evaluation given the constraints of the project handed to them and the BMJ for publishing a paper which whilst potentially confounded by a wide range of variables which might explain the null result, should be of interest to the taxpayer, who unwittingly funded the pilots and to health commissioners who, contrary to Ms Casey’s remarks, have a duty to purchase evidence-based interventions for their local population. Yours sincerely David Oliver [1] Guardian Unlimited. July 26th 2006. (Accessed 25/11/06) [2] The King’s Fund. Commentary on the primary care white paper “Your Health, Your Care, Your Say” Available at http://kingsfund.org.uk/resources/briefings (Accessed 25/11/06) [3] Department of Health, 2005The . National service framework for long
term conditions. Available at
[4] HealthCare Commission, Audit Commission, Commission for Social Care
and Inspectorate. Living well in later life. April 2005. Available at
[5] Kane R in lecture at British Geriatrics Society Spring Meeting, Harrogate, April 2005. [6] Kane RL, Keckhafer G, Flood S, Bershadsky B, Siadatty MS. The effect of Evercare on hospital use. J Am Geriatr Soc 2003; 51: 1427-54. [7] Singh D. Transforming Chronic Care. Evidence about people with long
term conditions. University of Birmingham Health Services Management
Centre. 2005. Available at
[8] Hutt R, Rosen R, McCauley J. Case-managing long term conditions. What impact does it have in the treatment of older people? London: Kings Fund. 2004. [9] National Library for Health. Primary Care Information Service. Answer
to question. “What is the best & most recent evidence for case
management in primary care as a means of reducing admission rates to
hospital?”. 17th August 2005. Available at
[10] Yuen P. BMJ readers should be even more cautious than usual in interpreting the results of the Evercare Evaluation. BMJ rapid responses 15 Nov 2006. [11] Williams NK. Redesigning primary care to reduce hospital admissions. BMJ rapid responses 15 November 2006. [12] United Health Europe. Assessment of the Evercare Programme in England
2003-4. Executive summary. February 2005. Available at
[13] Craig D 2005. Plundering the Public Sector. Competing interests: None declared |
|||
|
|
|||
|
Anthony P Roberts, Clinical Effectiveness Advisor North Tees PCT and South Tees Hospitals NHS Trust, TS17 6SF, Lucy Harding
Send response to journal:
|
Oliver mentions 10 PCTs in the Evercare pilot, for readers wondering where the 10th PCT not covered by the evaluation by Gravelle was, I can identify it as North Tees PCT where we piloted development of an NHS version of the data systems used by Evercare to identify patients suitable for case management ref to previous rapid response1. Our experience in the pilot helped us think through the development of Community Matron roles, identification of case-loads and monitoring and evaluation of their impact. Although the data tool we developed with UHE is no longer in use (substantially because of the loss of key staff), it remains the most integrated and powerful data tool we’ve yet used in the NHS. As members of the pilot group of PCTs we attended the meetings and watched with interest as the other PCTs grappled with the multiple problems involved in setting up the pilots. Evaluation was always going to be difficult and Gravelle have done as well as can be expected. Whilst a randomised trial obviously would be the strongest design for this (or other) interventions, it is disingenuous to claim that this could easily have been attained then (and was spurned for political reasons), or that it is likely to be easy to set up now. RCTs in this kind of intervention are difficult, expensive and time consuming with major methodological problems to overcome (for example, should a case management trial be cluster randomised or not?). Jenner makes many good points and similarly to their pilot we found the keys to (apparent) success revolve around the integration of case management with existing care for vulnerable elderly patients. This can be difficult to achieve particularly where case managers are trying to work in both primary and hospital care sectors, not least because secondary care doctors (even those still in training) can occasionally be aggressive and confrontational when presented with an unfamiliar staff group not under their control. There appears agreement that the goal of improving care for Very High Intensity Users is laudable – the quality and cost of care for these people is such that improvement must remain a key objective. Wherever evidence can be used to inform the development of policy and of services it should, but we should accept that it is unlikely to give clear, unambiguous answers and that a measured, reasoned approach will be needed. We continue to attempt to monitor and evaluate the Community Matrons we have in place in our PCT although no doubt we will not produce data that would satisfy everyone. Tony Roberts Head of Clinical Effectiveness, North Tees PCT and Clinical Effectiveness Advisor, South Tees Hospitals NHS Trust Lucy Harding Clinical Effectiveness Facilitator, North Tees PCT 1 Rapid Responses to: Martin Roland, Mark Dusheiko, Hugh Gravelle, and Stuart Parker Follow up of people aged 65 and over with a history of emergency admissions: analysis of routine admission data BMJ 2005; 330: 289-292: Evaluating Evercare nurses and data guided case management Anthony P Roberts, Alison Roe Competing interests: North Tees PCT was involved in the Evercare pilot. |
|||
|
|
|||
|
Martin J Howard, Service Improvement Manager King Square House, King Square, Bristol BS2 8EE
Send response to journal:
|
I wonder if, lost in the detailed and possibly valid argument about research methodology, the point of the work has been missed? In 2003 case-management was not a new concept, but combining case- finding and case-management was rare in older people’s care in the NHS. BMJ commentators are right to wonder whether the tail wagged the dog when policy on this approach emerged three years before the publication of the research from Roland et al. A few of us were on the inside of the ‘Evercare’ work. As we implemented the idea on the ground, nudging a new approach of holistic, active care into traditional ones about older people’s nursing, medical and social care, we were puzzled as to why the start of the Roland et al study was endlessly delayed by DoH mandarins, as if it was a political afterthought. However the emphasis on research is misleading. Local implementation at PCT level was never intended to be a research project. Instead, it was a bold local decision to ‘invest to save’ at a time when the Bristol health community was in financial crisis. Policy making is an illogical world, as BMJ correspondents might have noticed. Meanwhile, back in the real world, despite the research that tells us that it shouldn’t work, the approach does what we want it to: patients are alive and well and still living at home and avoiding hospital; and the PCT Board is pleased with the local evidence demonstrating that our Community Matrons more than cover their costs in emergency admissions saved [data triaged by a Nurse and Medical Consultant]. A final question. We have seen published in Hansard the investment costs of the United HealthCare Group contract with the DoH for the project. Have we yet seen figures on the cost-effectiveness of the commissioned research in improving lives? Competing interests: The author is an NHS employee, and has no financial or other connection with United HealthCare Group or its subsidiaries. |
|||
|
|
|||
|
David Oliver, Senior Lecturer, Elderly Care Medicine Institute of Health Sciences, University of Reading, RG1 5AQ
Send response to journal:
|
Sir I do not dispute for one second the views of both Howard [1] and Roberts [2] that proactive, joined up case management for frail elderly people with multiple long term conditions is a worthwile goal, nor that it would be an improvement on the fragemented crisis response that we often see now, nor that there are many good news stories from centres piloting case management. But that is not my point. And it doesnt do to try and paint commentators who express valid criticism as being ipso facto opposed to useful service developments. Simply asserting that something is working doesnt make it so. And it is simply not true that case management or other complex interventions in healthcare cannot be subjected to rigorous evaluation. Interventions such as Comprehensive Geriatric Assessment, or Stroke Unit care have been clearly demonstrated to be effective and other technologies (e.g. hospital at home for many conditions) been discredited using rigorous, independent evidence. It seems that the DH are committed to evidence based practice in some areas (e.g. NICE) and in others such as the Evercare pilot, happy to announce success or policy targets before the evidence is there - or worse still pretend that the evidence is already unequivocal. I am sorry, but this is what is really "disingenuous." [sic] The King's fund review [3] in 2004 had already identified several trials (mostly null) of case management as a technology before the Evercare Pilot - so they clearly can be performed. I believe there is a serious responsibility to spend public money first on ensuring that evidence based interventions are received by all those patients likely to benefit before spending money on interventions so lacking an evidence base. Even Nancy Williams from United HealthCare [4] has stated in her rapid response here that far more fundamental system redesign is required for case management to be effective in the NHS [4. As the authors of the Evercare evaluation stated [5] "Evercare introduced a variety of new services in primary care, with no resulting impact on hospital admission or bed utilisation". Do we believe the results of a multicentre evaluation performed by researchers with no vested interests or assertions by enthusiasts? [1] Howard MJ. BMJ rapid responses. 30/11/06 [2] Roberts AP. BMJ rapid responses 30/11/06 [3] Hutt R, Rosen R. Case Management in Long Term Care. Kings Fund 2004. [4] Williams NK. BMJ rapid response 17/11/06 [5] Gravelle, Dusheiko et al. BMJ 16/11/06 Competing interests: None declared |
|||
|
|
|||
|
david leopold, Consulatnt Physician Swansea, SA3 4JE
Send response to journal:
|
Dear Sir. United Health's responses to their latest flop are risible: rhetoric rather than evidence. Such evidence as exists about their schemes all points in the same direction. I'm sure that the hands-on NHS experts in this field could have designed an intervention which would demonstrably work. I'd guess the major reason for these repeated failures is founded in this: the pattern of hospital-using illness is predominantly acute, on a background of chronic disability, with a tendency to acute relapses. And accurate early and comprehensive multidisciplinary diagnosis and treatment is effective in preventing the strong tendency for illness to progress rapidly via impairment, to established disability, which is hard and time consuming to rehab. The worst gap in the Evercare model is diagnostic: even with many years of Consultant experience many illnesses pose a real challenge, and many require advanced diagnostic tests available quickly. This is the lesson from the 50 years UK development of geriatric Medicine, and until recently, the UK had the world lead in delivering cost -effective health care in this field in particular. A model which fails to achieve an accurate and timely diagnosis of the reason for a change in health status is doomed to fail. Kind regards David Competing interests: None declared |
|||
|
|
|||
|
Nicole L Klynman, GP registrar 18 North Hill, Colchester CO1 1DZ, Ugo Okoli, Director of Public Health, Enfield Primary Care Trust
Send response to journal:
|
Editor – Gravelle et al [1] undertook a useful evaluation of the impact of case management (Evercare) on elderly patients but failed to explore the reasons why the results of the NHS pilots were so different from results in the USA. The USA and UK healthcare systems as well as the case management programmes are significantly different which undoubtedly impacts on outcomes[2]. The US model employs more staff, many of whom are experienced medical practitioners, doctors as well as nurses, and case managers who are trained to degree level. Its staff work exclusively for the profit- making company and thus have a vested interest in the outcomes. The UK model uses nurses and case managers, not doctors. The nurses have little prescribing experience, often having recently completed a prescribing course. The american model also aims to support older people in nursing homes rather than the UK model of predominantly supporting people in their own home Research has also shown from studies of Kaiser Permanente that they achieved better performance at roughly the same cost as the NHS[3] with better systems integration, higher staffing levels[4], more efficient hospital management, competition and investment in information technology. Why are we allowing US companies to adapt models to the NHS when we already know they are unlikely to work? Surely we should be supporting UK based studies tailored to the ways of working of the NHS. The Castlefield study[5], reported in a non-peer reviewed journal is quoted as an example of UK success with case management. Surely we should be developing home- grown models that could work within our existing infra-structure rather than involving US corporations keen to have a financial interest in the UK. 1. Gravelle H, Dusheiko M, Sheaff R, Sargent P, Boaden R, Pickard S, Parker S, Roland R. Impact of case management (Evercare) on frail elderly patients: controlled before and after analysis of quantitative outcome data. BMJ doi:10.1136/bmj.39020.413310.55 (15 November 2006) 2. Hudson, B. Sea change or quick fix? Policy on long-term conditions in England. Health and Social Care in the Community. 13(4), 378-385 3. Feachem R, Sekhri N, White K. Getting more for their dollar: a comparison of the NHS with California's Kaiser Permanente. BMJ doi:10.1136/bmj.324.7330.135) 4. Ham, C. Learning from Kaiser Permanente: a progress report. http://www.natpact.nhs.uk/uploads/Kaiser%20Report%2023%20April%20.doc. 5. Castlefields Heath Centre. Chronic Disease Management. 2004 http://www.natpact.nhs.uk/eventmanager/uploads/castlefields.ppt Competing interests: None declared |
|||
|
|
|||
|
Louise Gibson, Senior Lecturer University of Chester, community matron students, University of Chester
Send response to journal:
|
Dear Sir Gravelle et al’s article has stimulated an interesting debate and this discussion has raised some pertinent points. However, it is the negative connotations of such research and the responses following its publication that could have a negative impact on patient care and outcomes. We acknowledge that one of the key targets of the Evercare (EC) model is to reduce hospital admissions but why was this failing not analysed in more detail? Also, the EC model has evolved since 2003 and although 2 hospital admissions is used as an indicator it is not the first indicator for case management referral as patients are referred through differing channels. The responses talk about the failings of the EC model but it is not discussed or considered in the wider policy/political context. We are now working in such a target driven way that any other achievements which does not have a code or clear defined outcome is deemed null and void. Consideration needs to be given to patient expectations and quality of life issues not just a reduction in hospital admissions and ticking of boxes. Nobody argues that there is a need to develop a body of evidence to support practice but the negative impact that this research has on the nurses who are working as community matrons is profound. Let’s not throw the baby out with the bath water and reject this model and approach to care. We need honest constructive evidence presented in the context of the bigger health policy picture and this needs to be debated. It is only when this is achieved will we be able to ascertain the outcome of the Evercare model and the role of the community matron. Kind Regards Community Matron Students
Ref: Hugh Gravelle, Mark Dusheiko, Rod Sheaff, Penny Sargent, Ruth Boaden, Susan Pickard, Stuart Parker, and Martin Roland. Impact of case management (Evercare) on frail elderly patients: controlled before and after analysis of quantitative outcome data BMJ, Nov 2006; doi:10.1136/bmj.39020.413310.55 Competing interests: None declared |
|||
|
|
|||
|
Andrew Clegg, SpR Geriatric Medicine Yorkshire Deanery
Send response to journal:
|
There are points in the response by Louise Gibson (1) that I would fully agree with. I echo her sentiments that the NHS is inappropriately hell-bent on hitting financial and clinically inappropriate targets. However, these are the yardsticks by which we are all now judged and if an intervention is set up specifically to meet certain targets, and is failing to achieve these targets (appropriate or not), we are obliged to reappraise the role of the intervention. I fully acknowledge the implications of the results of the Gravelle study (2) on the community matron body. However, Miss Gibson must also appreciate the present situation in hospital medicine. Funds are being incessently cut, wards are being closed and jobs are being lost on the basis of the community case management scheme. If the intervention that has been implemented to reduce readmission to hospital is failing, and therefore creating a persistent increase in the bed pressure in already overstretched hospitals we are, once again, obliged to reappraise the role of the intervention. I find it difficult to be sentimental about this. I agree that we need honest and constructive evidence that is presented in the context of the bigger healthcare picture but would argue that this is exactly what the study by Gravelle et al (2) provides. Unless we are provided with independent, unbiased evidence that the Evercare project is working in this country we are left with the inescapable conclusion that it is failing. Regards 1. Louise Gibson. Glass half full. BMJ Rapid Response, 7th December 2006 2. Hugh Gravelle, Mark Dusheiko, Rod Sheaff, Penny Sargent, Ruth Boaden, Susan Pickard, Stuart Parker, and Martin Roland. Impact of case management (Evercare) on frail elderly patients: controlled before and after analysis of quantitative outcome data BMJ, Nov 2006; doi:10.1136/bmj.39020.413310.55 Competing interests: SpR training in Geriatric Medicine |
|||
|
|
|||
|
Rowan H Harwood, consultant geriatrian, professor in geriatric medicine Nottingham NG5 1PB
Send response to journal:
|
Community case management ought to work. It is a good idea, and one which has been doing the rounds for a decade or more. In Nottingham, for many years, we had secondary care-based 'community interface nurses' who managed a case load of vulnerable ex-inpatients. They were inspired by the model of community psychiatric nurses, working across health service boundaries (developed by Tom Arie, a psychiatrist professor in an academic health care of the elderly department). The scheme worked well, but of course was not formally evaluated. Ironically, being hospital-based, this system folded when the imperative became to based care in the community, and funding was withdrawn in 2000. The risk is that by failing to evaluate robustly we lose the whole idea. The UHE model was always speculative rather than evidence-based. Sometimes you have to make the case to do an RCT using less-than-perfect observational studies. Gravelle et al bent over backwards to make something positive of their results, and they neither under- nor over- state their findings. But unlike defibrillation, or penicillin for meningitis, this turns out not to be an intervention which is so evidently superior that it can be accepted without an RCT. So that is what we need. There is other evidence. Konrad Jamrozik et al (Age and Ageing 2005) performed a cohort analysis of a pioneer case management scheme in West London. Hospital use was increased. In the late 1990s the Australian government funded a series of randomised controlled trials of community case management schemes. The results demonstrated no benefits, but, to their credit, they redesigned the intervention and commissioned a new series of trials. A further issue is to ask what we want community case management to achieve? Avoiding medical crises is laudable, but stating this as a desire to reduce emergency medical admissions is risky, and potentially age- discriminatory. Done well, the front end of hospital medicine is impressive, including for older people. This is the evidence-based gold standard (Andreas Stuck's Meta-analysis Lancet 1993). It is the tail of delayed discharged and inappropriately prolonged hospital stays that get hospital admission a bad name (and usually not the fault of secondary care). It does not follow that reducing admissions improves health. If case management can actually deliver more, more appropriate, more effective and more satisfying intervention that is a good thing, and is potentially open to investigation in a properly designed trial. But doubtless UHE will now be keen to fund the independent RCT they say is needed. There are a number of academic departments with the will and ability to deliver this, but it also requires honesty from policy makers (they do not know what actually works, and cannot know without empirical research) and a matching commitment to research. Competing interests: None declared |
|||
|
|
|||
|
david oliver, senior lecturer, geriatric medicine institute of health sciences, university of reading, london road campus, RG1 5AQ
Send response to journal:
|
Sir I completely endorse everything Prof Harwood has said, including - again, the need for better evidence on the radical restructuring of services - however intuitively appealing the idea of case management may be to practitioners and policy makers; and the fact that as multidisiciplinary comprehensive geriatric assessment has been well proven, we should not assume that emergency attendance at hospital represents a failure of care or is bad for the patient. Yet, in all the correspondance on these pages since the online publication of the Evercare Evaluation report 5 simple questions remain unanswered. Yes, there are plenty more, but can anyone help with these? I would love to hear some answers. And I note that both United HealthCare Europe and the representatives from the Department of Health have been silent since the initial correspondance. 1. Why did the DH pick Evercare when the overwhelming balance of the trial evidence at the time was against case management and when Evercare was very specific to US care home residents? And why did it not commission one or more RCTs or one or more type of case management instead of giving £4m straight to the private sector? 2. Why have the "Castlefields Data" - which claimed gains such as 15% reductions in re-admissions, 41% reductions in bed days, 31% reductions in length of stay, no social service overspend, no additional spend in primary care etc never been published in a peer reviewed journal yet heavily pushed by the DH despite the fact that no one else has replicated results of this magnitude in the UK. 3. Where on earth did the initial DH targets of 12% reduction in emergency bed days by 2008 (excluding children and mental health) come from? Or the statement that "each community matron would start with 50 patients on the caseload and build to 80"? There seems to be little basis in research of prior service models for either of these targets. 4. Why did so many statements and documents emanating from the DH give the impression that case management and the community matron role was already a sure fire winner? 5. Would United HealthCare have protested quite so loudly if the results of the evaluation by Gravelle et al had been positive? I think the answer to question 5 is self evident. The problem with the other four is that what happened led to wildly optimistic estimates of the degree and speed of benefit from case management - hence local so called "invest to save" cases built on bricks of straw and a faith that the advent of community matrons would solve a host of whole system problems. As HL Mencken said, "for every problem there is a solution which is simple, obvious and wrong" David Oliver Competing interests: None declared |
|||
|
|
|||
|
Wendy A M Walker, Senior Project Manager Luton PCT, LU1 1JD
Send response to journal:
|
The important debate over whether an Evercare type service can or cannot reduce unplanned hospital admissions will influence primary care trusts on whether to invest further in case management. In Luton teaching Primary Care Trust, one of the original Evercare pilots, we have developed the programme hugely and are confident that it is reducing unplanned hospital admissions. Indeed, we think that the service is currently saving around 900 unplanned admissions a year—a saving of around £3m at a cost of just under £1m. We thus plan to invest further in the service. The pilot in Luton ran from April 2003 to October 2004 and included initially four and then seven advanced primary practitioners supported by a pharmacist. Initially, the practitioners were attached to a specific general practice, and we used the original Evercare criteria for deciding which patients should be admitted to the service but soon became aware that both the process of care and the informatics needed to be refined to ensure a constant flow of the most acute patients with Long Term Conditions, this was also noted in the King’s Fund paper authored by Richard Lewis. Because we have shown the service is effective in both improving patient care and reducing unplanned admissions we have expanded and developed the service since the end of the pilot. We now have 14 advanced primary practitioners available seven days a week, managing a caseload of 500 patients. The practitioners now cover GP practices across the whole trust, and work closely with rapid response teams, district nurses, and the community rehabilitation teams. A bespoke database has been developed to capture accurate and informative data, and the criteria for admission to the programme has been developed accordingly, including the use of tools such as PARR and MIDAS to identify patients at risk of readmission to hospital. Information is collected on all patients, and a multidisciplinary team assesses when admissions have been adverted. The data collected on all admissions/aversions indicates a direct correlation between increases in adverted admissions and decreased hospital admissions. Perhaps the most convincing evidence that the programme is reducing emergency admissions is that admissions to the local district general hospital fell from just over 4100 in 2004-5 to just over 3800 in 2005-2006. Our neighbouring primary care trust saw a 10% increase in unplanned admissions. It’s important to note that it is only in the last year, 2 years since the start of the service that we are able to see a noticeable decrease in the number of unplanned admissions into hospital, reflecting the lead-in time for staff training and patient identification. This type of service is not a ‘quick fix’ it needs to be planned, supported and managed to achieve it’s potential and have a significant impact. Following it’s success, we plan with our commissioners to invest further in the service by increasing the number of advanced primary practitioners to cover a wider caseload, develop the skill mix within the team, integrate further with community teams and build further our information technology capacity. Wendy Walker
Competing interests: None declared |
|||
|
|
|||
|
john sharvill, GP Deal Kent ct14 7au
Send response to journal:
|
The latest article on the debate makes interesting reading especially as it showed (non statistical but large) increases in admissions and deaths. The first rapid repsonse from Wendy Walker does show what seems to be achievable but two things need expanding. Firstly she does not declare a conflict when it appears that she may have significant role in running this service. More importantly she says there is a saving of £3 million for a cost of £1m. Can she confirm how this £3 was spent or are we talking yet again about a paper figure? Did trust and PCT costs go down by a net £2m or was there an extra spending of £1m for this service with and a net extra spend overall? So often people talk of cost savings (as with statins) but total costs are increased. Competing interests: Our practice is involved in practice based comissioning and may be asked re similar projects unless central policy over-rule |
|||
|
|
|||
|
S Kapoor M.D., Res. Physician University of Illinois at Chicago, Chicago, IL, USA - 60612
Send response to journal:
|
The article “Impact of case management (Evercare) on frail elderly patients: controlled before and after analysis of quantitative outcome data” by Gravelle et al (1) was highly interesting. It was surprising to note that case management of elderly people via the evercare programme in Britain was not associated with a reduction in hospital admissions. This is surprising since the evercare programme has produced exactly opposite results across the atlantic in the US. Kane et al in a US study showed that the incidence of hospitalizations in ever care residents was nearly half the incidence of hospitalizations in control residents (2.43 versus 4.63). (2) Besides the average number of emergency department visits was also nearly 50 % less in evercare patients compared to the control group (3.37 per 100 enrollees per month versus 6.28 per 100 enrollees per month in the control group). Their study also showed that the evercare programme resulted in savings of nearly 100,000 dollars in hospital costs per nurse practitioner. The only explanation for these totally opposite results is the absence of intensive home nursing care for patients when they became ill in the UK plan, unlike the US plan where domiciliary nursing care for sick patients is a part and parcel of the evercare programme. Clearly the incorporation of this feature in the UK plan can go a long way in reducing emergency room visits and hospital admissions. Other changes that might decrease hospital admissions include closer follow up of patients after hospital discharge thus preventing readmissions. (3) Also more extensive training of nurse practitioners to distinguish high risk patients from low risk patients can reduce admission rates. Finally, preferential recruitment of nurse practitioners specializing in geriatric care may go a long way in decreasing hospital admission rates.(4) 1. Gravelle H, Dusheiko M, Sheaff R, Sargent P, Boaden R, Pickard S, et al. Impact of case management (Evercare) on frail elderly patients: controlled before and after analysis of quantitative outcome data. BMJ 2007; Jan 6;334(7583):31. 2. Kane RL, Keckhafer G, Flood S, Bershadsky B, Siadaty MS. The effect of Evercare on hospital use. J Am Geriatr Soc 2003; Oct;51(10):1427 -34. 3. Naylor M, Brooten D, Jones R, Lavizzo-Mourey R, Mezey M, Pauly M. Comprehensive discharge planning for the hospitalized elderly. A randomized clinical trial. Ann Intern Med 1994; Jun 15;120(12):999-1006. 4. Burl JB, Bonner A, Rao M, Khan AM. Geriatric nurse practitioners in long-term care: demonstration of effectiveness in managed care. J Am Geriatr Soc 1998; Apr;46(4):506-10. Competing interests: None declared |
|||
|
|
|||
|
Martin Roland, Director National Primary Care Research and Development Centre, University of Manchester, M13 9PL, Hugh Gravelle, Mark Dusheiko, Rod Sheaff, Penny Sargent, Ruth Boaden, Susan Pickard, Stuart Parker.
Send response to journal:
|
We would like to correct some misconceptions by Peter Yuen (1) about our paper on the evaluation of Evercare. First, we did correct for baseline differences between intervention and control practices - in two ways. We compared the change in outcomes before and during the intervention in intervention practices with the change in outcomes for the control practices. We also used propensity score matching to match intervention and control practices. The assertion that our follow up period was only 8 months is also incorrect. We reported results for a post intervention period (October 2004 to March 2005) which is 12 to 18 months after the median patient was enrolled. In relation to the length of exposure, correcting for the number of patients enrolled in each practice made no difference to the results. We also allowed for potential contamination of control groups by excluding practices in PCTs where we were aware of similar interventions. This left around 7000 control practices. Our evaluation included an extensive qualitative analysis which addresses many of the comments made by other correspondents about this paper. This will be published separately, but is also available on the web (2). Mr Yuen points to the inherent weaknesses of observational studies. We agree, but also note that the study on which UnitedHealth's claim of a 50% reduction in admissions (for a different intervention) was based was also a case-control study. We fully agree with other correspondents that in an ideal world, randomised controlled trials should be undertaken before introducing policies with potentially large health and cost consequences. This is especially important where the literature suggests that the effects of an intervention depend heavily on the context in which it is introduced. 1. Yuen P. BMJ readers should be even more cautious than usual in interpreting the results of the Evercare evaluation. www.bmj.com/cgi/eletters/334/7583/31#149297. 2. NPCRDC. National Evaluation of Evercare. www.npcrdc.ac.uk/r5.37. A synopsis of lessons learned from Evercare for the future of case management in the NHS will also be published in the Health Service Journal on 1st March Competing interests: Authors' response |
|||
|
|
|||
|
Carmel M Martin, Associate Professor of Family Medicine Northern Ontario School of Medicine, Canada K1N 5E3, Joachim P Sturmberg
Send response to journal:
|
This is a very important case study (1) demonstrating the unintended consequences resulting from health service decision makers implementing ‘simple’ and/or ‘complicated’ interventions in a health system that is ‘complex’. (2) The buying an ‘off the shelf’ solution to the care management and coordination needs of frail, ill older patients seems like a good idea – all that is required is some local process adaptation and the product will deliver the desired outcomes anywhere, as long as it has been demonstrated the intervention is effective. Yet in the case of Evercare, the implementation did not deliver the intended outcomes. (1) A rush of reflection follows, culminating with the authors’ conclusions: “in an ideal world, randomised controlled trials should be undertaken before introducing policies with potentially large health and cost consequences”. (3) In fact, it appears that according to these methodological criteria (which originally were designed to apply to simple technologies such as lithotripsy machines), not only this evaluation study, but also the original Evercare study were not properly evaluated. Furthermore, most of the existing evidence has limited external generalisability, because “the literature suggests that the effects of an intervention depend heavily on the context in which it is introduced.”(3) Thus it doesn’t matter how many randomized controlled trials (RCT) or other study types are conducted, their findings will lack external generalisability because of widely varying health service contexts and different actors. However, such studies are useful, in as much as a high level policy synthesis of the findings indicate that the principles of case management are valid and can meet the needs for complex care for frail ill patients. Success is based on local context, local actors and the mode of implementation. Using a complexity approach, once the principles are accepted as valid and desirable to introduce to an organization, the most relevant research is to develop an understanding (often through qualitative evaluation) about what worked, what didn’t, and where and why etc.. In fact, this reflects a major component of the response by the authors. (3) So where does that leave us? Is it rationale to buy an ‘off the shelf’ product for health service reform like one would buy a vacuum cleaner and by merely using an adapter plug, have a product that works in many different countries and settings? Of course not! Case management is an intervention in a health care system based on principles of care coordination and management with the aim to improve quality of care and life for an at risk population. (4) However a health care system behaves like a complex adaptive system, it best organizes itself through a ‘bottom up’ approach. Hence what matters in terms of implementation of a case management system is: clear and valid policy and principles; local ownership and buy in; assessment of the health care system capacity and functionality to respond to the needs of frail people; and customization to address the local community characteristics. (5) To this end, interestingly, the first round of Australian Coordinated Care trials demonstrated these key points. Every one of the 9 local trials experimented with different locally determined processes to meet community needs for care coordination. The RCTs proved a nightmare in methodological terms because there was such diversity in capacity, needs and implementation in each local setting. RCT outcomes evaluation was problematic despite a core data set and central rules across the trials. The major learnings were the about the processes of local needs assessment, understanding of the how local systems worked and how they might be improved in future. (4) We are left with compelling knowledge - both tacit and explicit - that we ought to look at health systems as complex adaptive wholes with an ‘in-built’ property to adapt and emerge in multi ways to the same stimulus. Understanding different systems adaptation to case management policy and principles and how to transfer such knowledge to other settings should be the aim of research. Complex systems are organic and respond to planting seeds and nurturing existing strengths rather than the imposition of obvious ‘simple’ top down solutions.(2) However time is needed to allow the system to adapt and grow in new directions.(5) Carmel M Martin and Joachim P Sturmberg carmel.martin@NorMed.ca References (1) Gravelle H, Dusheiko M, Sheaff R, Sargent P, Boaden R, Pickard S, et al. Impact of case management (Evercare) on frail elderly patients: controlled before and after analysis of quantitative outcome data. BMJ 2007; Jan 6;334(7583):31. (2) Glouberman, S 2001 Towards a New Perspective on Health Policy: Final Report CPRN Study No. H|03 www.healthandeverything.org (3) Martin Roland, et al.Authors' response bmj.com, 31 Jan 2007 (4) Esterman AJ, Ben-Tovim DI. The Australian coordinated care trials: success or failure? The second round of trials may provide more answers. Med J Aust. 2002 Nov 4;177(9):469-70. (5) Sturmberg JP and Martin C. Primary Care reforms – a complex adaptive system in Sturmberg JP The Foundations of Primary Care. Daring to be different. Radcliffe Medical Press 2007. Competing interests: None declared |
|||