Frede Olesen Former GP Professor, DrMedSci Research Unit for General Practice, Aarhus University
Frede Olesen Some of my first observations 1976 - 85
§ Variation in the use of hospital-based A&E services ranged between 3,5% and 35% per inhab. per year in Danish counties § Aarhus County: average 15%. A small city 15 km from centre used 7%, and of these 50% could have been treated in primary care § (my first conversation with a very angry consultant …) § Use of intermediate beds in a municipality based nursing home § a substantial reduction in number of waiting days in hospital beds before discharge to a nursing home § (a very angry leading geriatric consultant …) § A national survey showed no correlation between number of nursing home places/inhab. and number of waiting days before discharge from hospital
Frede Olesen The consequence of the first angry consultant … Data create political awareness
1983: A report from The national Board of Health Variation from 4% to 37% between Danish counties with respect to use of A&E hospital service
Frede Olesen A & E service in emergency rooms London Kings-College hospital § Three trial groups based on staff at first contact § A senior hospital doctor – A young hospital doctor – A GP § Inclusion: patients with primary care problems (41%, n= 4641) § Method: random allocation of patients – follow-up with questionnaires to patiens and GPs § Outcome: no difference in satisfaction, number of follow up contacts within 10 days. § Costs of treatment:
Direct costs Costs including referrals
Senior Dr.
Young Dr.
GP
Dalet et al. BMJ 1996;312: 1340-4
Frede Olesen Contacts to the out of hours service 1990-2012
% Telephone consultations Reform 58%
Clinic consultations 32%
Home visits 10%
Mill 3 2.9 No. of contacts
2,5
2
Morten Bondo Christensen & Linda Hüiberts Research Unit for General Practice, Aarhus
Frede Olesen The new OoH service in Copenhagen 2013
§ We changed triage from GPs to nurses § can anybody be surprised that major changes happened in triage results? § number of 112 calls and hospital referrals increased substantially § We changed consultations from GPs to hospital pediatricians § can anybody be surprised that number of hospital admissions increased substantially
Frede Olesen Variation in hospital bed days for childhood astma (blue) and use of ambulatory astma care visits (red) in 13 Danish counties
Moth G &al. UFL 2006;168:56-61
Frede Olesen % of practice pop. as frequent attenders
Frequent attender = belongs to upper 10% of attenders in a county. N=255 GPs.
Average number of freguent attenders per GP: 10%
Frede Olesen Vedsted et al. 1998 Number of consultations in the GP surgeries in day time in 268 GP settings in Aarhus county 2003 X-axis: 268 GPs ranged according to activity. Y-axis number of consultations 90% variation: 1,7
800
700
600
500
400
300
200
100
0
Kilde: Vedsted et al. Almen lægepraksis i Danmark, 2005
Frede Olesen A health care system has different boundaries - a small change can have major consequences
Hospital 11 0,81
GP 1010 9
Self care
90 91
Frede Olesen Risk of a cardiovascular event among T2 diabetics in more than170 GP settings Adjusted for age, sex, previous cardiovascular disease, cancer and previous prescription of cholesterol lowering drugs before diagnosis by opportunistic screening
Odds ratio 1 1.7 (1.0-3.1) 3.1 (1.6-6.0) 2.6 (1.3-5.2) Intervention: 1 2.5 (1.2-5.6) 2.9 (1.2-7.0) 2.0 (0.8-5.6) Control: 1 0.5 (0.2-1.4) 3.9 (1.5-10.1) 3.0 (1.2-7.6)
11 intensive 15 routine practices had no patients treated with cholesterol lowering drugs
Intensive treatment practices Routine practices RK Simmons, AH Carlsen, S Griffin et al. Diabet Med. 2014 Sep 3. doi: 10.1111/dme.12574.
Frede Olesen www.addition.au.dk – slide lent from prof. Torsten Lauritzen PROMS and PREMS in cancer follow-up
§ PROMS: patient recorded outcome measures § PREMS: patient recorded experience measures
§ A tool to enable patient centred follow-up and meet patients’ needs
Frede Olesen PROM – a communication tool for instance in targeted follow-up for prostate cancer
Follow-up Follow-up Follow-up 3 months 6 months 12 months
L
Frede Olesen Liv Dørflinger, The Danish Cancer society Frede Olesen § Assessment: § Blood sample + PROM A B C
Letter Telephone Outpatient (doctor/nurse) appointment (doctor/nurse)
Frede Olesen Liv Dørflinger, The Danish Cancer society What is your need for a consultation in hospital? (n=42)
52%
26% 14% 7% I will call myself if I Telephone Outpatient Don't know need a consultation consultation appointment Frede Olesen Liv Dørflinger, The Danish Cancer society International variations raise questions: palliative care and outgoing teams § More than 80 % prefer to die at home
§ The Netherlands: 30% die in hospital § Denmark: 55% die in hospital § US: large differnces amonguniversity hospitals
§ Denmark: hospital-basedoutgoing palliative teams § The Netherlands: focuson GPs and local district nurses
Frede Olesen Frede Olesen Evidence and health care delivery § Effective care:
§ Advantages much bigger than disadvantages § Everybody in the guideline target group should be treated § Variation = underconsumption = bad performance § Preference-sensitive care:
§ There are different treatment options(for instance: operation/prescription/wait and see) § Information of the patient and shared decision-making are essential § Variation sensitive to doctors preferences and doctor-paternalism § Supply-sensitive care : § Supply and access will increase consumption § For instance number of follow-up visits. Interval between follow -up. Available medical technology
Frede Olesen Some observations and interpretations - what influences variation?? § Organisation of service … (A&E – palliative care) § Culture (A&E – palliative care) § Distance (A&E) § Physician speciality and skills (A&E London) § Physician-beliefs about good or bad care (asthma, palliative care) § Understanding of PROMs and patients’ needs (prost. cancer) § Physician-competence, knowledge or personal beliefs (cholesterol) § Incentives (pay for performance in GP, DRG-groups, municipalities) § A lot of blind spots among physicians and decision makers § Lack of understanding of Wennberg’s 3 groups
Frede Olesen Problem number 1 - education and knowledge among health care planners
§ Health care planners and a growing number of administrators do not understand that wants and needs are as dynamic as they are § the most ignorant group think that all needs are objective § They do not understand even simple determinants of wants and needs § Too many believe in simple marketmechanisms § the example with Danish municipality remunerationof regions § Nearly none of those who really understand somethingabout wants and needs know Wennberg’s groups § look for instance at the debate about prioritiesin medicine
Frede Olesen Problem number 2 - problems with the doctor community
§ Lack of understandingof the same problems as in previous slide § Lack of interest in health services research § Insufficient knowledgeabout the dynamics across the boundaries of differentparts of health care § A specialDanish problem?? § any problem: § could be solved if we had more staff belonging to my own profession § could be solved if we had more money and staff in our prof./dept.
Frede Olesen Some of the problems in summary § Education in the determinants for use of health service is minimal § Understanding of Wennberg’s 3 groups is not explicit § Lack of awareness of an ever ongoing upgading of most initiatives and activities in W-groups 2 and 3 followed by regular cost cuts in all 3 groups. Has led to quality problems in group 1. § Incentive and budget problems and failures § Too poor medical leadership in health care – an educational gap § Lack of willingness to collaborate between administrative and medical leaders § Too few look at or analyse health care as a comprehensive system § Phycisians do not take quality deviations seriously enough
Frede Olesen Are modern challenges reflected in the quality and research agenda? - we need much more research in ways to deliver good health care
Frede Olesen And the angry physicians …
When we came to know each other I discovered that they were very nice and polite people … And I think they discovered that I was not as ignorant as they thought … They were angry because they were professionally ambitious and simply because they did not know about variation and reasons for variation This is the real challenge for graduate and postgraduate education
Frede Olesen Conclusions § The fact: variation will continue to exist § The good news: variation may lead to good new developments § The problems and challenges: § The professional medical and administrative understandingof the nature of variation must be improvedsubstantially § All adminstrative staff and medical staff must spend much more time on this book: § A new education for administrative staff candidatingfor leadershippositions in healthcare – a new administrative educational line in medical specialisation for doctors § Muchmore health services research is needed
Frede Olesen