How to match supply and demand?

1. How to match supply and demand? Bertrand Guidet Medical Intensive Care Unit Paris , France

2. Supply ? • Physicians • Registered nurses (RN) • Helpers • Head nurse • Other personnel • Respiratory therapist • Pharmacist • Clerks • Psychologist

3. Demand ? • Patient treatment • Administrative tasks • Research • Teaching • Quality assessment?

4. Specificities ? • Type of hospital • Type of ICU • Case mix • Seasons • Crisis situation • Winter • Terrorism • Pandemia

5. Countries specificities ? • Working rules • ICU as part of a broader department • Flexibility? • Supplementary working hours • Autonomy in decision making? • Pool of nurses available in the hospital? • Interim agencies?

6. Shortage in trained physicians in ICU?

7. Why fearing shortage in Europe? • In the present time : • The trained ICU physicians are getting older • There is a reduction in the working time • 48 h a week • No clinical activity after on night duty during a 11 h period • In the future : • Reduction in the number of physicians • Reluctance to choose CCM as a specialty : • To much work including nights duty • Medico-legal issue

8. Is the question so simple? • 1- What is an ICU ? • 2- What is a trained ICU physician ? • 3- What are ICU physicians supposed to do ? Job, tasks, duty, ….. • 4- What is the impact of the shortage ? • 5- Recommandations

9. What is an ICU ? • AnOfficial text has defined ICU in France « décret » April 5th 2002 • Minimum requirements : - Number of beds : at least 8 - ICU director certified in intensive care medicine - Physician dedicated solely to the ICU during the night - Non medical personnel : * Patients to nurses ratio : 2.5 / 1 * Patients to nurses’aides ratio : 4 / 1

10. Other criteria • Ability to provide : • Continuous monitoring • Organ support : mechanical ventilation, dialysis techniques, cardiovascular support,…. • Proximity to other hospital units • Emergency room • Operating room • Radiology department • Description of the activity • Severity scores • Workload indexes • Case mix

11. According to these criteria,at least 30% of existing ICUwill close in France in the next 5 years

12. What is the ICU’s contribution to the hospital activity ? • Besides diagnosis and treatment of patients admitted to the ICU • To perform procedures for patients not admitted in ICU • Insertion of central venous line • Dialysis • Broncho alveolar lavage or transbronchial biopsy for hypoxic patients. • To evaluate patients for ICU admission • Ethical issue  Clinical activity is not restricted to the patients admitted in the ICU

13. What is a trained ICU physician ? Teaching of critical care in Europe The results of 2 surveys • Garcia-Barbero. Crit Care Med 1996, 24 : 696 • No standardisation of curriculum content • No clear definition of competence (knowledge, attitudes, skills, and judgement) necessary to practice • little coordination of postgraduate training • Bion. Intensive care Med 1998, 24 : 372. • Formal training in ICM : 18/21 countries (85%) • Length of training : 18 to 30 months (median 24 months) • Access to ICM specialty • Multidisciplinary : 57% • Anaesthesia (28%) • Accreditation in ICM : 18 countries • including 12/18 with dual registration in a base specialty and in ICM

14. Partial list of trained physician activities • Coordination of patient care • Continuous quality improvement • Committees • Morbidity and mortality review • Risk management • Safety net • Education • Medical consultant both formal and informal • Medical-legal issues • Policy maker • Epidemiology and infection control • Resource allocation (includes triage) • Research • Conflict resolution • Technology assessment • Computers/Medical information system

15. And… • Health care policy • Interhospital relations and planning • Social services and coordination • Liaison with • Hospital administration • Physician staff • Nursing staff • Respiratory therapist • Laboratory • Radiology • Department of medicine, surgery, anaesthesia • Clergy:pastoral care • Self education • Research • Teaching • ……..

16. Schedule of French ICU physiciansone week survey in 2002, 32 units including 13 teaching hospital

17. Organisational models of ICUs • Organisational model : • Open units : patients remains in the service of the admitting physician • Closed units : medical director and designee screen all admissions and discharges and assume direct patient care responsibilities • Closed units : • Reduction of LOS, morbidity and mortality • Requires more doctors

18. Example of intensive communication intervention Lilly, Am J Med 2000; 109 : 469 • Method : multidisciplinary meetings held within 72 hours of critical care admission : patients, families, critical care team. + follow-up meetings to discuss palliative care options when continued advanced supportive technology could not achieve the patient’s goal. • The implementation of this active communication has • reduced the median LOS (4 vs 3 days; p = 0.01) • while the mortality remained the same or even decreased (31% vs 23%; p = 0.06).

19. Impact of organisational characteristics of ICU to outcomePronovost , JAMA 1999, 281 : 1310 • Example of abdominal aortic surgery. • Maryland hospitals; 1994-1996 • Measure of the impact of not having daily rounds

20. Effect of a medical intensivist on patient careMathous Mayo Clin Proc 1997, 72 : 391

21. Factors that increase ICU resource use after abdominal aortic surgery Pronovost , JAMA 1999, 281 : 1310 • Not having daily rounds • Having a nurse-patient ratio < 1:2 • Not having monthly review of mortality and morbidity • Extubating patients in the operating room

22. Relation between physician staffing and performance • Most of the studies have focused on the nurses and very few have looked at the impact of the number or qualifications of the physicians on outcomes. • What is performance ? • Mortality • Morbidity • LOS • Patient and family satisfaction • Other physician satisfaction • Administration • ……. • Methodological limitations : • Type of hospital and environment • Case mix • Admission and discharge policy

23. In practice What is the reality ? What are the proposals ?

24. Observations from the department of Veterans affairs’ ICUHalpern Crit care Med 1994, 22 : 2008 MICU SICU Combined ICU director Director assigned (%) 100 90 83 CCM training (%) 33 26 21 CCM board (%) 52 37 38 ICU attendings coverage 24h (%) 93 89 71 attending rounds (%) 91 76 60 Fellows assigned (%) 86 18 20

25. Medical staff of French ICUa 1999 survey of 174 units Type of hospital Non teaching Teaching ICU (n) 130 44 Beds (n) 10.3 19.9 LOS (days) 6.6 8.3 Mechanical ventilation (%) 53 55 Full time physician (n) 2.7 3.3 Part time (n) 0.1 0.7 Fellows (n) 0.8 1.8 Full time equivalent (n) 3.6 5.8 FTE / bed 0.35 0.29 Residents (n) 0.9 3.1

26. Night duties Type of hospital Non teaching Teaching Type of night duties (%) On site (%) 91.5 100 For ICU solely (%) 71 84 Together with resident always (%) 14 47 sometimes (%) 16 32 never (%) 70 21 On calls (%) 27 34 Physicians on the list (%) n 7.6 10.5 % of Dr belonging to the ICU 59 66 Qualification of physicians (%) Medical doctors 95 88 CCM specialists 77 44

27. Critical care services and personnel : recommendations based on a system of categorization into two levels of careAmerican College of Critical Care Medicine of the SCCM. Crit Care Med 1999, 27 : 422 • Medical staff organisation • A distinct medical staff • The team is organized and led by an intensivist • Patient management is directed by an attending physician who : • Has clinical management responsibility • Is board certified in CCM • Sees the patients as often as required but at least twice daily • Participation in the institution’s bioethical committee

28. Physician availability 24-hr in-house coverage • Non tertiary center : At least one physician who can manage emergencies. If this requirement is fulfilled by senior residents, an attending physician fully credentialed in CCM must be on call and available within 30 mins. • Tertiary center : Critical care physician is appropriately credentialed to provide dedicated care to the critical care unit patients: If this requirement is fulfilled by critical care fellows, a critical care staff physician must be on call and available within 30 mins.

29. The French model • Maximum working time per week : 48 h, including on nights duties. • Rest of at least 11 hours after on night duty • Minimum requirements for a 10 beds unit : • 3 physicians for morning • 2 physicians for afternoon • 1 physician for night  This requires 6.5 FTE per unit

30. Conclusion : unresolved issues • Impact of restructuration of the units • Impact of creation of intermediate care units • New organisation and management of the ICU • Increase the attractiveness of this specialty • Ethical issue – admission policy • The expected shortage of ICU physicians will increase the risk of lower quality of care • No team building • Lack of coherency in the clinical management of patients • No involvement outside the unit • No research and teaching

31. Recommendation for RN? • Stratification in three levels of care • Level 1 : 1 RN for 1 patient • Level 2 : 1 RN for 2 patients • Level 3 : 1 RN for 4 patients • In France and at the bed side • 1 RN for 2.5 patients • 1 Helper for 4 patients

32. Adaptation according to work load • Measure of work load • Tiss, PRN, NEMS,… • Inclusion of new tasks • Paper work • Communication • Quality assessment

33. Time frame? • Short range • Phone call 2 hours prior to the working shift? • Long range • Discrepancies between supply and demand

34. Short range • Ability to mobilize RN • Same unit • Other units • Stop admission for this shift

35. Long range • Definition of norms and standard • Country level • European recommendations? • Type of norms : Quantitative & qualitative • Ratio RN/patient • Basic training of RN • Continuous education

36. What happen if these recommendations are not fulfilled? • Reduction of the number of beds • Impact for the non admitted patients? • Regional regulation • Number of ICU beds/100,000 inhabitants • Increase of work load for each individual RN • Impact on quality of care • Nosocomial infection • Burn out and turn over • Weaning procedures

37. Analysis of the work flow according to type of ICU • Surgical ICU with scheduled surgery • Less RN in morning shift • Less RN during week end • Tertiary center • Polytrauma center

38. Is it possible to predict activity?

39. MODELS FOR FORECASTING THE NUMBER OF EMERGENCY DEPARTMENT VISITS Wargon M, …. Guidet B. Emergency Med J (in press). • We reviewed articles retrieved by a Medline search for studies of models designed to predict patient attendance in EDs or walk-in clinics. • Only 9 studies were identified. • Most of the models used to predict patient volume were either linear regression models including calendar variables or time series models. These models explained 31% to 75% of patient-volume variability. • Although day of the week had the strongest effect on patient volume, this variable explained only part of the variability, whose causes remained largely unidentified. Adding meteorological data failed to improve model performance.

40. Numer of patients attending one specific ED : Real figures

41. Predicted versus observed ED visits in 4 hospitals

42. Identification of variables influencing ED visits Hospital R2 Day of week Official holiday months school holidays N°1 0.461 p= 0.000 p=0.280 p=0.008 p= 0. 394 N°2 0.433 p=0.081 p=0.003 p=0.000 p=0.0712 N°3 0.471 p= 0.001 p= 0.012 p=0.000 p=0.002 N°4 0.433 p= 0.000 p= 0.551 p = 0.02 p=0.033 All 4 EDs 0.631 p= 0.000 p= 0.005 p = 0.00 p=0.019

43. 1. Originalsignal: observed patient visits Training set Validation set 3. model evaluation 2. estimation of the parametervalues in the model Forecasted set with the model Training set Performance evaluation Best fit Validation set Construction and validation of the model Model Parameter values

44. Organization of intensive care units, in case of pandemic avian flu. Guery B, Guidet BRev Mal Respir. 2008;25: 223-35. • Working hypothesis • ICU should be expended twice its capacity • Only one third of the nurses will at work • Sick themselves • Children sick • Transportation difficulties • Nursery closed • Schools closed • High work load per patient • High incidence of ARDS • Individual protection in order to prevent contamination

45. Proposal in case of pandemic avian flu. Guery B, Guidet BRev Mal Respir. 2008;25: 223-35. • Reduction of scheduled hospital activities in particular for surgery enabling mobilization of personnel working in this sector to the ICU • Allowing nurses without expertise in ICU to work with ICU nurses with a ratio 1/1. • Dedicated personnel to handle communication with the relatives, the administration, the media… • Ethical issues • Admission policy • Decision to withdraw or withheld treatment

46. Importance of management skills and organisation Relation between organizational score and work loadAssessing organizational performance in ICU’s: A French experience. Minvielle E, …, Guidet B. J Crit Care 2008, 23:236-244

47. Methods • 26 ICUs located in the Paris area, France • Data were collected through answers of 1000 ICU personnel to COMIC questionnaire and from the database. • Organizational Performance was assessed through a composite score related to five dimensions: • Coordination and adaptation to uncertainty, • Communication, • Conflict Management, • Organizational change and Organizational Learning, • Skills developed in relationship with patients and their families.

48. ORGANIZATIONAL PERFORMANCE • Organizational learning and change • Communication • Coordination • Problem-solving/conflict management • Skills developed in the relation patient/caregivers

49. The effect of Individual and ICU level factors on Organizational performance using Hierarchical Modelling.

50. Relations between perceived workload-burnout and performanceEuricus I study (n=2009 questionnaires) -.19*** PWL EE DP PUP .54*** -.17*** .70*** PWL : perceived workload EE : emotional exhaustion DP : depersonalization PUP : perceived unit performance