- Email: [email protected]
The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units
G Model
ACCPM-27; No. of Pages 7 Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
Original article
The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units Jean-Yves Lefrant a,*, Bernard Garrigues b, Ce´line Pribil c, Isabelle Bardoulat d, Fre´de´ric Courtial d, Fre´de´rique Maurel d, Jean-E´tienne Bazin e, the CRREA Study Group1 with the collaboration of the AzuRea Group a
Division anesthe´sie re´animation douleur urgences, faculte´ de me´decine, universite´ Montpellier 1, CHU de Nıˆmes, place du Professeur-Robert-Debre´, 30029 Nıˆmes cedex 9, France b Service de re´animation et de surveillances me´dico-chirurgicales polyvalentes, centre hospitalier du Pays d’Aix, Aix-en-Provence, France c Health Outcomes Department, GlaxoSmithKline, 100, route de Versailles, 78163 Marly-le-Roi cedex, France d IMS Health, Health Economics and Outcomes Research Department, Tour Ariane, 5-7, place de la Pyramide, 92088 La De´fense cedex, France e Service anesthe´sie re´animation, CHU de Clermont-Ferrand, 1, place Lucile-Aubrac, 63003 Clermont-Ferrand cedex, France
A R T I C L E I N F O
A B S T R A C T
Article history: Received 20 December 2013 Accepted 1st September 2014 Available online xxx
Objectives: To estimate the daily cost of intensive care unit (ICU) stays via micro-costing. Methods: A multicentre, prospective, observational, cost analysis study was carried out among 21 out of 23 French ICUs randomly selected from French National Hospitals. Each ICU randomly enrolled 5 admitted adult patients with a simplified acute physiology II score 15 and with at least one major intensive care medical procedure. All health-care human resources used by each patient over a 24-hour period were recorded, as well as all medications, laboratory analyses, investigations, tests, consumables and administrative expenses. All resource costs were estimated from the hospital’s perspective (reference year 2009) based on unitary cost data. Results: One hundred and four patients were included (mean age: 62.3 14.9 years, mean SAPS II: 51.5 16.1, mean SOFA on the study day: 6.9 4.3). Over 24 hours, 29 to 186 interventions per patient were performed by different caregivers, leading to a mean total time spent for patient care of 13:32 05:00 h. The total daily cost per patient was s1425 s520 (95% CI = s1323 to s1526). ICU human resources represented 43% of total daily cost. Patient-dependent expenses (s842 s521) represented 59% of the total daily cost. The total daily cost was correlated with the daily SOFA score (r = 0.271, P = 0.006) and the bedsidetime given by caregivers (r = 0.716, P < 0.0001). Conclusion: The average cost of one day of ICU care in French National Hospitals is strongly correlated with the duration of bedside-care carried out by human resources. ß 2015 Published by Elsevier Masson SAS on behalf of the Socie´te´ franc¸aise d’anesthe´sie et de re´animation (Sfar).
Keywords: ICU Cost
1. Introduction * Corresponding author. Tel.: +33 4 66 68 30 50. E-mail address: [email protected] (J.-Y. Lefrant). 1 CRREA Study Group: Dr Ali Mofredj, centre hospitalier de Salon; Dr Marc Boudon, centre hospitalier Dreux; Dr Luc Jacques, centre hospitalier du Bassin de Thau; Dr Anne Marie Guerin-Robardey, centre hospitalier de Beauvais; Dr Jean Cabalion, centre hospitalier de Selestat; Dr Christian Roth, CHI Annemasse Bonneville; Dr Olivier Delastre, CHI Elbeuf Louviers; Dr Andre´ Dupont, centre hospitalier de Fontainebleau; Dr Didier Demory – Dr Jacques Durand-Gasselin, CHI Toulon; Dr Pierre Cougot, CHU Toulouse; Dr De´sire´ Samba, CHU Coˆte de Nacre Caen; Dr Jean Pierre Quenot, CHU de Dijon; Dr Jean Marie Tonnelier, CHU de la Cavale Blanche; Dr Boris Jung et Dr Ge´rald Chanques, CHU Saint-E´loi, Montpellier; Dr Daniel Villiers, CHU de Nantes; Dr E´ric Lobjoie, CHU d’Amiens; Dr de Filippis, clinique mutualiste chirurgicale; Dr Emmanuel Cantais, CHU Marseille Timone; Dr Jean Pierre Fulgenci, CHU Tenon, Paris; Pr Jean Paul Mira, hoˆpital Cochin, Paris.
Patients with life-threatening disease and/or organ failure(s) are classically admitted to Intensive Care Units (ICU). According to the type of ICU and patient disease, the mortality rate ranges from nearly 5% to 60% [1]. In addition to the treatment of causative pathology, the management of ICU patients involves the use of organ support therapies such as mechanical ventilation (invasive or non-invasive), vasoactive drug infusion and renal replacement therapy. The use of these organ support therapies can vary among different types of ICUs [2]. In the USA, ICU beds, also including intermediate care, account for less than 10% of hospital beds and 22% of total hospital costs
http://dx.doi.org/10.1016/j.accpm.2014.09.004 2352-5568/ß 2015 Published by Elsevier Masson SAS on behalf of the Socie´te´ franc¸aise d’anesthe´sie et de re´animation (Sfar).
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004
G Model
ACCPM-27; No. of Pages 7 J.-Y. Lefrant et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
2
[3]. In 2005, ICU beds represented nearly 15% of hospital beds and the national cost of ICU stays was estimated at $81.3 billion, which represented 13.4% of hospital care, 4.1% of national health expenditures and 0.66% of the gross domestic product [4]. A German study involving 51 ICUs reported that total ICU costs represented 20% of overall hospital costs, 62% of which was associated with human resources [5]. In France, hospital/private clinic-related expenses represent more than 44% of all national health-care costs. Although ICU beds represent less than 5% of hospital beds in France (mainly in public hospitals), they consume around 10% of hospital costs [6]. Despite this large cost contribution, published estimates of the daily cost of an ICU stay vary considerably due to the heterogeneity of the methods used for its assessment and the profile of considered patients [7–10]. Therefore, the main objective of the CRREA (Couˆt re´el d’une journe´e en re´animation: actual cost of one day in the ICU) study was to estimate the actual daily cost of adult ICU stays in France, whatever the patient’s profile and the time of admission, using a microcosting approach.
2. Patients and methods The present study (GSK protocol 113382) was strictly observational and was approved by the French Society of Anaesthesiology and Intensive Care (Socie´te´ franc¸aise d’anesthe´sie et de re´animation [Sfar]) and the French Intensive Care Society (Socie´te´ de re´animation de langue franc¸aise [SRLF]). According to French law (law 88-1138 relative to Biomedical Research of December 20th 1988 modified on August 9th 2004), this noninterventional study did not require approval by an Ethics Committee nor informed, signed consent from patients. The study was declared to, and approved by, the National Committee for Data Processing and Civil Liberties (authorization # 1338734 obtained on March 20th 2009). However, under French law, all patients or their representatives were informed of the study by the ICU physician and could decline participation to the study. 2.1. The study design The present study was designed by a scientific committee composed of 3 intensivists (JYL, BG, JEB), 3 health economists (CP, FM, FC) and one statistician (IB) and followed 4 successive phases. 2.1.1. Phase 1: site selection and recruitment In order to well represent the overall pool of French ICUs, the present study required the participation of 23 ICUs: 12 university hospitals (centre hospitalier universitaire [CHU]), 10 general hospitals (centre hospitalier [CH]), and 1 private, non-profit, public service hospital (e´tablissements participant au service public hospitalier [PSPH]). The 3 ICUs (2 CHUs and 1 CH) managed by the 3 intensivists of the scientific committee participated in the study. The remaining ICUs were randomly selected from the French National Hospital database and stratified by hospital category (CHU, CH or PSPH). The head manager of each randomised ICU was contacted to ask his/her willingness to participate in the study. If and when he/she refused, a new randomly selected ICU was contacted until 20 ICU head managers accepted participation (Fig. 1). 2.1.2. Phase 2: patient enrolment In each ICU, 5 randomly selected adult patients were included within a one-month period between May 10th and October 16th 2009. In 22 units, only one patient per study day was included, thus resulting in 5 different day-waves over which the study was carried out. In one unit, all five patients were included on the
Fig. 1. Flow chart describing the inclusion process of ICUs into the study. BMI: body mass index; CH: centre hospitalier (French General Hospital); CHU: centre hospitalier universitaire (French University Hospital); ICU: intensive care unit; PSPH: e´tablissements participant au service public hospitalier (French private, non-profit, public service hospital); RRT: renal replacement therapy; SAPS II: simplified acute physiology score II; SOFA: sequential organ dysfunction score.
same day. Patients were included regardless of the day of the week. Patient inclusion criteria (according to the French definition of ICU patients [decree April 5th 2002]): adult patients ( 18 years of age) admitted to an ICU with a simplified acute physiology score (SAPS II) [11] 15 and requiring at least one major ICU medical procedure in accordance with the French hospital procedure-based pricing (corresponding to organ failure support; Table 1) could be included into the study. Non inclusion criteria: Patients with an expected life expectancy < 24 h, patients with a life-sustaining decision at the beginning of the daily assessment, patients for whom an ICU discharge was planned within a 24 hour period and patients previously enrolled in the present study did not participate.
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004
G Model
ACCPM-27; No. of Pages 7 J.-Y. Lefrant et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx Table 1 Medical acts in accordance with the French hospital procedure-based pricing. Requirement of infusion of dopamine or dobutamine or noradrenalin or adrenalin Mechanical ventilation 3 days Mechanical ventilation with FiO2 > 0.6 and/or PEP > 6 cmH2O Fluid loading > 50 mL/kg/24 h within one day Non-invasive ventilation > 12 h per day Renal replacement therapy for acute renal failure Red blood cell pack transfusion > ½ blood volume Fresh frozen and/or platelet transfusion for dessiminated intravascular coagulation Intracranial pressure monitoring Cardiopulmonary resuscitation Pericardial drainage External electric shock in emergency Intra-aortic conter-pulsion Mechanical ventilation in separated lungs Plasma exchange Electro-systolic stimulation probe insertion
2.1.3. Phase 3: data collection for a 24-hour period for each patient Activity data: in each ICU, the numbers of patients and different caregivers in the Unit were recorded on the study day. Patient data: age, gender, height, weight, co-morbidities, reason for admission to the unit, and SAPS II score at admission were collected [11]. The number of organ failures on the study day (Severity Organ Failure Assessment [SOFA] score [12]) was assessed. Time of care dedicated to each patient (indicated as hours:minutes [hh:min]): All hospital staff (regardless of speciality) taking part in patient management were asked to self-time the duration of care they carried out for the patient management of each intervention. These data were divided into direct (for instance, bedside-care), indirect (for instance, the time a doctor spends informing relatives of the patient’s status, and/or syringe preparation by nurses) and administrative (for instance, patient report preparation by secretaries) interventions. When an included patient was discharged from the Unit (death or transfer) during the study day, data were collected and extrapolated for a 24-hour period. Medications, materials, biological and radiological exams used for the included patient’s care were recorded. 2.1.4. Phase 4: economic assessment phase The prices of all resources identified were estimated from the hospital’s perspective (reference year 2009) based on the cost data provided by each centre.
(A) Staff time dedicated to patient care: for each hospital staff category, the cumulative time dedicated to patient care (both direct and indirect tasks) was multiplied by the average hourly wage for that staff category. (B) Other caregiver dependent times: certain indirect tasks are affected by the number of patients present in an ICU, and the time dedicated to these tasks is calculated in accordance with the proportion occupation of the ICU. These tasks included Night Watch Standby time (excluding intervention time), day standby time (excluding intervention time), the completion of transversal patient verifications during respite, cross cutting time watch, nosocomial infection prevention meetings, etc. (C) Medications: the costs of all given treatments were estimated using prices provided by each centre and the resources used by each patient. (D) Big consumables (such as haemodialysis devices, etc.) were valued using unit prices provided by each centre and the resources used by the patient. (E) The daily cost of all small consumables (such as bandages, gauzes, Renal Replacement Therapy [RRT] materials, etc.) was
3
extrapolated from the total cost of the ICU and the number of days completed. (F) The costs of laboratory tests and other investigations were directly calculated using the biological laboratory test and procedures classification (Nomenclature des actes de biologie me´dicale [NABM]; class B for biological) and the unit cost of these biological tests in each hospital. (G) The costs of other investigations were directly calculated using the common classification of medical procedures (Classification commune des actes me´dicaux [CCAM]) and the cost of these procedures in each hospital. (H) ICU administrative expenses were divided into structural costs, amortizing expenses, general logistic expenses and medical logistics. All such data were collected by the on-site study coordinator in collaboration with the financial and administrative departments. The estimation of an average patient-day cost in the ICU was based on information collected at the site level and taking into account the annual number of patient-days both at the site and ICU levels. Using these previous definitions, the actual daily cost could be separated into human resource allocation (A + B), patient-dependent expenses (A + C + D + E + F + G) and administrative expenses (B + H). 2.2. Statistical analysis Quantitative variables are expressed as mean values with standard deviations (SD) and median values with inter-quartiles [Q1–Q3] according to the distribution of the studied parameter. In order to better express studied parameters, the ranges (lowest value-highest value) were added. Qualitative variables are expressed as frequencies with percentages. The total daily cost was expressed with its 95% confidence interval. When adequate, a Pearson’s correlation coefficient was calculated. All analyses were performed using SAS system software version 9.2 (SAS Institute Inc., Cary, NC, USA) under WindowsTM. A 2-sided type 1 error rate of 0.05 was considered as the threshold for statistical significance. 3. Results 3.1. Intensive Care Units During the study period, 23 ICUs initially accepted to participate in the study, but one CHU did not include any patients. Therefore, the present study included 109 patients (one CH included 4 patients) (Fig. 1). In one CHU, the medical staff time was not assessed and the associated 5 patients were excluded for per protocol analysis, thus resulting in 104 patients in 21 ICUs (10 CHUs, 10 CHs and 1 PSPH) (Fig. 1). Measures of hospital activity in 2008 are shown in Table 2.
Table 2 Measures of activity (2008) for the studied ICUs (median values with range). Number of authorized ICU beds (n = 21) Number of opened ICU beds (n = 19) Number of admitted patients (n = 21) Number of ICU days (n = 21) Bed occupancy (%) (n = 19) Mean length of stay in ICU (days) (n = 21) Patients requiring mechanical ventilation (%)(n = 17) Patients requiring mechanical ventilation > 48 h (%)(n = 20) Mortality rate (%)(n = 18)
10 10 425 3478 100 7.2 65
(6–20) (8–18) (131–988) (1830–6900) (81–100) (5.1–20.6) (44–96)
43 (17–96) 21 (12–35)
For each studied parameter, the number of valid responses is given in parentheses.
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004
G Model
ACCPM-27; No. of Pages 7 J.-Y. Lefrant et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
4
Table 4 Patient severity.
3.2. Patients Patients (67 males, 64%) characteristics and severity are shown in Tables 3 and 4. Mean age was 62.3 14.9 years. At admission, the mean SAPS II score was 51.5 16.1 and the mean body mass index was 28.3 11.0. The median time elapsed from admission to the study day was 5.8 days [1.8–19.5]. On the study day, the mean SOFA score was 6.9 4.3. Ninety patients (87%) were mechanically ventilated, 55 (53%) were sedated and 17 (16%) required renal replacement therapy. There were 4 patients with withholding (n = 2) and withdrawing (n = 2) decisions. Two patients died without lifesustaining decisions, whereas 3 patients were prematurely discharged from the ICU. Patient characteristics were similar when the analysis included 22 intentions to treat ICUs (109 patients) (data not shown). 3.3. Staff time Over 24 hours, 29 to 186 interventions per patient were performed by different caregivers leading to a mean total time Table 3 Patient characteristics. Age (years) 62.3 14.9 65 [53–73] Sex ratio (F/M) 38/66 2 Body Mass Index (kg/m ) (n = 94) 28.3 11.0 26.5 [23–30.8] Medical history Patients with at least one 87 clinically relevant medical history 2.6 1.92 [1–4] Number of following clinically relevant medical histories 48 Hypertensiona Risk of multiresistant bacteria 38 b carriage 29 Sedentarity Diabetes mellitus 21 Tobacco abusec 21 Obliterant arteriopathy 16 Dyslipidemia 16 Coronary artery disease 15 d Chronic renal failure 15 e Alcohol abuse 15 New York Heart Association 10 (NYHA) > III 10 Chronic respiratory deficiencyf Stroke (ischemic or hemorrhagic) 5 Gastro-duodenal ulcer 5 Clinical hepatopathy 4 Main diagnosis Severe sepsis and/or septic shock 28 Acute respiratory failure 22 Haemorrhagic shock 13 Other shocks 7 Coma 6 Immediate postoperative surveillance 5 Acute exacerbation of Chronic 5 Obstructive Pulmonary Disease 5 Voluntary medicine intoxication Status epilepticus 3 Acute pancreatitis 2 Stroke 2 Miscellaneous 6
(18–87) (37%/63%) (14–99.1) 84% (0–8) 46% 37% 28% 20% 20% 15% 15% 14% 14% 14% 10% 10% 5% 5% 4% 27% 21% 13% 7% 6% 5% 5% 5% 3% 2% 2% 6%
For each studied parameter, the number of valid responses (when different from 104) is given in parentheses. For continuous variables, the results are expressed in mean value with standard deviation, median value with [inter-quartiles] and (extreme values). For categorical variables, data are expressed in absolute values with (percentages). a Systolic arterial pressure > 140 mmHg and/or diastolic arterial pressure > 90 mmHg. b Patient who was treated with antibiotics or who was admitted to hospital in the previous 6 months. c > 20 boxes per year. d Basal creatinine > 150 mMol/l or renal replacement therapy. e Daily alcohol intake > 140 g (female) or > 210 g (male). f Requiring intermittent or continuous oxygenotherapy.
51.5 16.1 50 [40–61] SAPS II score at admission SOFA score at inclusion (n = 103) 6.9 4.3 6 [4–10] Time from ICU admission to study day 19.6 39.2 5.8 [1.8–19.5] Mechanical ventilation 90 Patients with sedative drugs (n = 103) 55 17 Patients with current renal replacement therapy (n = 103) Medical act in accordance with the French hospital procedure-based tariffs Number of medical acts per patient 2.5 1.3 2 [1–3] Dopamine or dobutamine > 8 mg/kg/min 59 or nor/epinephrine infusion 58 Mechanical ventilation > 2 days with PEP < 6 cmH2O and FiO2 < 0.6 Mechanical ventilation with PEP 43 6 cmH2O or FiO2 60% Fluid loading < 50 mL/kg/24 h within 31 one day 25 Renal replacement therapy Non-invasive ventilation 24 Cardiopulmonary resuscitation 7 2 External electric shock Red blood cell pack transfusion > ½ 2 blood volume 2 Fresh frozen and/or platelet transfusion for DICa 1 Intra-aortic conter-pulsion Intracranial pressure monitoring 1
(21–97) (0–19) (0–229) 87% 53% 17%
(1–6) 57% 56% 41% 38% 24% 22% 7% 2% 2% 2% 1% 1%
For each studied parameter, the number of valid responses (when different from 104) is given in parentheses. For continuous variables, the results are expressed in mean value with standard deviation, median value with [inter-quartiles] and (extreme values). For categorical variables, data are expressed in absolute values with (percentages). a Disseminated intravascular coagulation.
spent for patient care of 13:32 05:00 hours (range = 05:08– 31:48)(nurse time = 51%) (Fig. 2). The corresponding total cost was s350 s161 (Table 5). 3.4. Daily cost of ICU stays The total daily cost per patient was s1,425 s520 (Table 5) (95% CI = s1,323 to s1,526). The different components of cost are shown in Table 5 and in Fig. 3. The cost of ICU human resources represented 43% of total daily cost. Patient-dependent expenses (s842 s521) represented 59% of the total daily cost. There was a correlation (r = 0.977, P < 0.001) between patient-dependent expenses and the total daily cost. Patient-dependent expenses were also correlated with the daily SOFA score (r = 0.271, P = 0.006) and bedside-care time (r = 0.716, P < 0.0001), but not with the SAPS II score (r = 0.129, P = 0.191). In consequence, there was a correlation between the total daily cost and the SOFA score (r = 0.230, P = 0.019)(no correlation with admission SAPS II, r = 0.094, P = 0.341). Centre dependent expenses represented 41% of the total daily cost with no correlation with total daily cost (r = 0.103, P = 0.300).
4. Discussion In this prospective observational study including 104 patients in 21 French ICUs, micro-costing was performed in order to assess the daily cost of an ICU patient. The median value of the daily cost was s1425 (95% CI = s1323 to s1526), ranging up to s3173. The major component of this cost was ICU human resources (s607), reflecting in part the time spent at the bedside by caregivers (mean value = 13:32). The second highest component was the administrative expenses assigned to the ICU (s326). Other costs
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004
G Model
ACCPM-27; No. of Pages 7 J.-Y. Lefrant et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
Fig. 2. Number of caregiver interventions per patient-day and time (expressed in hours and minutes) spent for patient care.
related to patient-dependant activities (medications, consumables, analyses) were rarely over s130 per day. However, the use of certain supplies, like haemodialysis or hemofiltration and associated consumables could largely increase this cost component. More importantly, the total daily cost was strongly related to SOFA scores, thus suggesting a relationship with patient severity (and not the severity at patient admission) and time spent at the bedside by caregivers for patient care.
5
The strength of the present study resides in the use of a microcosting approach applied to 104 randomly chosen patients in 21 different ICUs randomly selected from a representative panel of French ICUs (taking into account the proportions of public/private, University and non-University hospitals). This methodology is extremely time consuming and difficult to carry out, and 48 ICUs declined participation, and 2 were excluded at the beginning of the study organization. However, micro-costing resulted in a complete study measuring the time spent by different caregivers for patient care, the assessment of medications, consumables, and analyses used. Furthermore, the administrative expenses associated with French ICUs were assessed. To our knowledge, though Elliott [6] proposed this type of methodology in 1997, no previous, prospective study has been performed in the ICU setting with this methodology. Our findings are concordant with those reported by Tan et al. [13] in a retrospective study involving 7 European ICUs. The authors of the latter reported a total daily cost for an ICU patient ranging from s1168 to s2025. However, laboratory, imaging and administrative costs and charges were not assessed. Moreover, the time spent by different caregivers was not measured. Halpern et al. [4] reported an increase in per day ICU costs, increasing from $2698 to $3518 (+30.4%) between 2000 and 2005. This difference could probably be explained by the different methodologies used in such studies [7–9]. Intensive care units are costly on a per-patient-day basis, even if ICU beds classically represent only about 10–15% of available hospital beds [3,4]. In the USA and Netherlands, the cost of an ICU has been reported to be close to 20% of the total hospital budget [3,4,14]. However, the daily cost estimation for an ICU varies from s855 in a multicentre German study [5] to s3221 in US ICUs [4]. Many factors may explain these differences between countries. The patient case-mix, the bed occupancy rate, the density of acute care beds, the definition, the type and the use of ICUs in each country and the staff composition are evidently a major part. However, medical practice, the availability of health-care resources, the hospital payment system (public/private-mix and insurance payment) and the relative and absolute costs between national systems could influence the ICU daily cost [7,14–16]. Cost differences between studies could also be the consequence of the different methodologies used for daily cost assessment rather than real differences in daily cost [7,10,16]. Finally, the precision of the different components of cost could be questioned [13]. Though we applied a complete micro-costing approach, some limitations should be noted. First, the present ICUs and patient samples may not be representative of ICUs and ICU patients.
Table 5 Total daily cost per patient and its different components in euros (s). Values are expressed as means SDs and in medians with [inter-quartiles] and (ranges). Parameters
Mean SD
Median, [IQ], (range)
Total daily cost Time dedicated to patient care (A) Other caregiver dependent times (B) Human resources (A) + (B) Medications (C) Consumables (D) Materials for RRT Materials for RRT/patient with RRT (n = 18) Laboratory tests (E) Investigation tests (F) Sub-total patient-dependent expenses (A) + (C) + (D) + (E) + (F) Administrative expenses assigned to ICU (G) Structural costs Amortization expenses General logistics expenses Medical logistics Sub-total centre dependent expenses (B) + (G)
1425 520 350 161 257 54 607 161 125 223 140 230 45 141 261 246 139 135 88 134 842 521
1263, [1062–1628], (684–3173) 310, [250–412], (100–1,080) 245, [224–270], (158–431) 580, [508–664], (315–1289) 57, [25–137], (0–1692) 60, [35–114], (13–1533) 0, [0–0], (0–728) 168, [50–396], (2–728) 104, [60–162], (0–778) 47, [0–86], (0–962) 668, [496–1021], (159–2729)
326 104 62 32 37 29 157 95 70 60 583 113
313, [267–355], (174–613) 64, [40–78], (14–133) 30, [19–41], (7–146) 168, [99–209], (8–400) 56, [40–68], (20–311) 554, [522–625], (384–833)
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004
G Model
ACCPM-27; No. of Pages 7 J.-Y. Lefrant et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
6
€ 400
40%
€ 350
€ 350
35%
€ 326 € 300
30%
€ 257
€ 250
25%
€ 200
20%
€ 150
€ 140
€ 100
15%
€ 139
€ 125
25%
23%
18%
10%
€ 88
€ 50
10%
5%
10%
9%
6%
€0
0%
Time dedicated Other caregiver Medications to patient care dependent time
Con sumables Laboratory tests Investigation tests
Contribution to the daily total cost
Administrative expenses assigned to ICU
Average daily cost 95% CI
Fig. 3. Distribution of the daily total cost of ICU stays per patient and average cost of each component per patient-day.
However, the scheduling of units and patients was randomised and took into account the proportion of private/public institutions and University and non-University institutions in France. Moreover, the panel of 104 patients involved in the present study was similar to patient characteristics reported by a recent French study focused on the activity of 66 ICUs in the South of France and involving 605 patients [17]. The French system of care and reimbursement could impair the extrapolation of our findings to other countries. However, our results agree with those of previous studies performed in different countries. At admission, patient severity is similar to those reported in recent European observational studies, such as the EPIC and SOAP studies [18–20]. The proportion of ventilated patients and the mortality rates in the studied ICUs are in the range of the most severe patients recorded in the recent assessment of ICU activity in the United States of America [1]. Secondly, the daily cost of an ICU patient could vary according to ICU staff composition [10]. In France, Critical Care is defined by law (decree April 5th, 2002) and pertains to patients in whom at least two different organ failures are present or could occur, in whom vital prognosis is engaged and in whom technical organ support is required. Under the same law, Intensive Care is defined by a unit managing patients with only one organ failure, whereas Intensive Surveillance manages high risk patients with no organ failure (potential risk of organ decompensation). Another law (edited on the same date) determined the human resources required in Critical Care Units; at least two nurses are required to manage 5 patients, whereas 1 aid nurse is required to manage 4 patients. In terms of reimbursement, an admitted patient with a SAPS II score 15 associated with at least one reimbursementqualifying procedure (medical acts corresponding to the artificial support of an organ failure) results in a reimbursement of s850 per day spent in the ICU. This payment is added to the French DRG tariff allocated to the patient hospital stay. This previous definition helps explain our choice of inclusion and exclusion criteria. This patient population corresponds to the most severe patients admitted in ICUs in other countries [1]. The wide range of daily costs reported in the present study could be also explained by the time elapsed from admission to the time of micro-costing. Indeed, Dasta et al. [21] demonstrated that the two first days in the ICU were the most expensive. The cost of direct patient care was closely correlated to SOFA scores and, in contrast, the total daily cost was correlated with
institution-dependent costs. This finding means that patient severity directly impacts daily costs. The most serious patients required more care time, resulting in an increase in the cost of direct patient care and the total daily cost. Tan et al reported similar findings in European ICUs [13]. The ratio patient/caregiver is classically considered as an indicator of ICU quality and has been shown to be associated with fewer nosocomial infections and better quality of care [22–26]. In conclusion, this prospective observational study using a micro-costing approach demonstrated that the total daily cost of an ICU patient ranges from s684 to s3173, and is largely influenced by direct patient care. Contributorship Jean-Yves Lefrant, Bernard Garrigues, Ce´line Pribil, Isabelle Bardoulat, Fre´de´ric Courtial, Fre´de´rique Maurel, Jean-E´tienne Bazin: conception and design, analysis and interpretation of data, drafting the article or revising it critically for important intellectual content, and final approval of the version to be published. All other members of CRREA Group: collecting data and analysis. Jean-Yves Lefrant, Bernard Garrigues and Jean-E´tienne Bazin were consulted in order to define the study design, data analysis and the initial written version of the article. Fre´de´rique Maurel, Fre´de´ric Courtial and Ce´line Pribil helped with the design, the organization of the study and data management. Isabelle Bardoulat performed data management and statistical analysis. Disclosure of interest Jean-Yves Lefrant, Bernard Garrigues, and Jean-Etienne Bazin served as paid consultants for GlaxoSmithKline. Ce´line Pribil is employed by GlaxoSmithKline. Isabelle Bardoulat, Fre´de´rique Maurel, and Fre´de´ric Courtial are employed by IMS Health. Acknowledgements The authors thank the medical and nursing staff of each unit participating in the study and the help of Carey Suehs (CHU Nıˆmes) for the writing of the article.
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004
G Model
ACCPM-27; No. of Pages 7 J.-Y. Lefrant et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
Funding: This study was supported by grants from GlaxoSmithKline and was contracted by IMS Health, France. References [1] Lilly CM, Zuckerman IH, Badawi O, Riker RR. Benchmark data from more than 240,000 adults that reflect the current practice of critical care in the United States. Chest 2011;140:1232–42. [2] Constantin JM, Leone M, Jaber S, Allaouchiche B, Orban JC, Cannesson M, et al. Quelle activite´ et quels personnels soignants dans 66 unite´s de re´animation du sud de la France? Ann Fr Anesth Reanim 2010;29:512–7. [3] Halpern NA, Bettes L, Greenstein R. Federal and nationwide intensive care units and healthcare costs: 1986-1992. Crit Care Med 1994;22:2001–7. [4] Halpern NA, Pastores SM. Critical care medicine in the United States 20002005: an analysis of bed numbers, occupancy rates, payer mix, and costs. Crit Care Med 2010;38:65–71. [5] Moerer O, Plock E, Mgbor U, Schmid A, Schneider H, Wischnewsky MB, et al. A German national prevalence study on the cost of intensive care: an evaluation from 51 intensive care units. Crit Care 2007;11:R69. [6] Fourrier F. Pre´sentation de la re´animation; 2011, http://www.srlf.org. [7] Elliott D. Costing intensive care services: a review of study methods, results and limitations. Aust Crit Care 1997;10:55–63. [8] Pines JM, Fager SS, Milzman DP. A review of costing methodologies in critical care studies. J Crit Care 2002;17:181–6. [9] Clement Nee Shrive FM, Ghali WA, Donaldson C, Manns BJ. The impact of using different costing methods on the results of an economic evaluation of cardiac care: microcosting vs gross-costing approaches. Health Econ 2009;18:377–88. [10] Reis Miranda D, Jegers M. Monitoring costs in the ICU: a search for a pertinent methodology. Acta Anaesthesiol Scand 2012;56:1104–13. [11] Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA 1993; 270:2957–63. [12] Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A, Bruining H, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. Intensive Care Med 1996;22:707–10.
7
[13] Tan SS, Bakker J, Hoogendoorn ME, Kapila A, Martin J, Pezzi A, et al. Direct cost analysis of intensive care unit stay in four European countries: applying a standardized costing methodology. Value Health 2012;15:81–6. [14] Oostenbrink JB, Buijs-Van der Woude T, van Agthoven M, Koopmanschap MA, Rutten FF. Unit costs of inpatient hospital days. Pharmacoeconomics 2003;21: 263–71. [15] Ridley S. Critical care outcomes. Anaesthesia 2001;56:1–3. [16] Edbrooke DL, Ridley SA, Hibbert CL, Corcoran M. Variations in expenditure between adult general intensive care units in the UK. Anaesthesia 2001;56: 208–16. [17] Leone M, Ragonnet B, Alonso S, Allaouchiche B, Constantin JM, Jaber S, et al. Variable compliance with clinical practice guidelines identified in a 1-day audit at 66 French adult intensive care units. Crit Care Med 2012;40:3189–95. [18] Vincent JL, Sakr Y, Sprung CL, Ranieri VM, Reinhart K, Gerlach H, et al. Sepsis in European intensive care units: results of the SOAP study. Crit Care Med 2006;34:344–53. [19] Vincent JL, Bihari DJ, Suter PM, Bruining HA, White J, Nicolas-Chanoin MH, et al. The prevalence of nosocomial infection in intensive care units in Europe. Results of the European Prevalence of Infection in Intensive Care (EPIC) Study. JAMA 1995;274:639–44. [20] Vincent JL, Suter P, Bihari D, Bruining H. Organization of intensive care units in Europe: lessons from the EPIC study. Intensive Care Med 1997;23:1181–4. [21] Dasta JF, McLaughlin TP, Mody SH, Piech CT. Daily cost of an intensive care unit day: the contribution of mechanical ventilation. Crit Care Med 2005;33: 1266–71. [22] Hugonnet S, Chevrolet JC, Pittet D. The effect of workload on infection risk in critically ill patients. Crit Care Med 2007;35:76–81. [23] Hugonnet S, Uckay I, Pittet D:. Staffing level: a determinant of late-onset ventilator-associated pneumonia. Crit Care 2007;11:R80. [24] de Vos M, Graafmans W, Keesman E, Westert G, van der Voort PH. Quality measurement at intensive care units: which indicators should we use? J Crit Care 2007;22:267–74. [25] Garretson S. Nurse to patient ratios in American health care. Nurs Stand 2004;19:33–7. [26] Blot SI, Serra ML, Koulenti D, Lisboa T, Deja M, Myrianthefs P, et al. Patient to nurse ratio and risk of ventilator-associated pneumonia in critically ill patients. Am J Crit Care 2011;20:e1–9.
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004
ACCPM-27; No. of Pages 7 Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
Original article
The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units Jean-Yves Lefrant a,*, Bernard Garrigues b, Ce´line Pribil c, Isabelle Bardoulat d, Fre´de´ric Courtial d, Fre´de´rique Maurel d, Jean-E´tienne Bazin e, the CRREA Study Group1 with the collaboration of the AzuRea Group a
Division anesthe´sie re´animation douleur urgences, faculte´ de me´decine, universite´ Montpellier 1, CHU de Nıˆmes, place du Professeur-Robert-Debre´, 30029 Nıˆmes cedex 9, France b Service de re´animation et de surveillances me´dico-chirurgicales polyvalentes, centre hospitalier du Pays d’Aix, Aix-en-Provence, France c Health Outcomes Department, GlaxoSmithKline, 100, route de Versailles, 78163 Marly-le-Roi cedex, France d IMS Health, Health Economics and Outcomes Research Department, Tour Ariane, 5-7, place de la Pyramide, 92088 La De´fense cedex, France e Service anesthe´sie re´animation, CHU de Clermont-Ferrand, 1, place Lucile-Aubrac, 63003 Clermont-Ferrand cedex, France
A R T I C L E I N F O
A B S T R A C T
Article history: Received 20 December 2013 Accepted 1st September 2014 Available online xxx
Objectives: To estimate the daily cost of intensive care unit (ICU) stays via micro-costing. Methods: A multicentre, prospective, observational, cost analysis study was carried out among 21 out of 23 French ICUs randomly selected from French National Hospitals. Each ICU randomly enrolled 5 admitted adult patients with a simplified acute physiology II score 15 and with at least one major intensive care medical procedure. All health-care human resources used by each patient over a 24-hour period were recorded, as well as all medications, laboratory analyses, investigations, tests, consumables and administrative expenses. All resource costs were estimated from the hospital’s perspective (reference year 2009) based on unitary cost data. Results: One hundred and four patients were included (mean age: 62.3 14.9 years, mean SAPS II: 51.5 16.1, mean SOFA on the study day: 6.9 4.3). Over 24 hours, 29 to 186 interventions per patient were performed by different caregivers, leading to a mean total time spent for patient care of 13:32 05:00 h. The total daily cost per patient was s1425 s520 (95% CI = s1323 to s1526). ICU human resources represented 43% of total daily cost. Patient-dependent expenses (s842 s521) represented 59% of the total daily cost. The total daily cost was correlated with the daily SOFA score (r = 0.271, P = 0.006) and the bedsidetime given by caregivers (r = 0.716, P < 0.0001). Conclusion: The average cost of one day of ICU care in French National Hospitals is strongly correlated with the duration of bedside-care carried out by human resources. ß 2015 Published by Elsevier Masson SAS on behalf of the Socie´te´ franc¸aise d’anesthe´sie et de re´animation (Sfar).
Keywords: ICU Cost
1. Introduction * Corresponding author. Tel.: +33 4 66 68 30 50. E-mail address: [email protected] (J.-Y. Lefrant). 1 CRREA Study Group: Dr Ali Mofredj, centre hospitalier de Salon; Dr Marc Boudon, centre hospitalier Dreux; Dr Luc Jacques, centre hospitalier du Bassin de Thau; Dr Anne Marie Guerin-Robardey, centre hospitalier de Beauvais; Dr Jean Cabalion, centre hospitalier de Selestat; Dr Christian Roth, CHI Annemasse Bonneville; Dr Olivier Delastre, CHI Elbeuf Louviers; Dr Andre´ Dupont, centre hospitalier de Fontainebleau; Dr Didier Demory – Dr Jacques Durand-Gasselin, CHI Toulon; Dr Pierre Cougot, CHU Toulouse; Dr De´sire´ Samba, CHU Coˆte de Nacre Caen; Dr Jean Pierre Quenot, CHU de Dijon; Dr Jean Marie Tonnelier, CHU de la Cavale Blanche; Dr Boris Jung et Dr Ge´rald Chanques, CHU Saint-E´loi, Montpellier; Dr Daniel Villiers, CHU de Nantes; Dr E´ric Lobjoie, CHU d’Amiens; Dr de Filippis, clinique mutualiste chirurgicale; Dr Emmanuel Cantais, CHU Marseille Timone; Dr Jean Pierre Fulgenci, CHU Tenon, Paris; Pr Jean Paul Mira, hoˆpital Cochin, Paris.
Patients with life-threatening disease and/or organ failure(s) are classically admitted to Intensive Care Units (ICU). According to the type of ICU and patient disease, the mortality rate ranges from nearly 5% to 60% [1]. In addition to the treatment of causative pathology, the management of ICU patients involves the use of organ support therapies such as mechanical ventilation (invasive or non-invasive), vasoactive drug infusion and renal replacement therapy. The use of these organ support therapies can vary among different types of ICUs [2]. In the USA, ICU beds, also including intermediate care, account for less than 10% of hospital beds and 22% of total hospital costs
http://dx.doi.org/10.1016/j.accpm.2014.09.004 2352-5568/ß 2015 Published by Elsevier Masson SAS on behalf of the Socie´te´ franc¸aise d’anesthe´sie et de re´animation (Sfar).
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004
G Model
ACCPM-27; No. of Pages 7 J.-Y. Lefrant et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
2
[3]. In 2005, ICU beds represented nearly 15% of hospital beds and the national cost of ICU stays was estimated at $81.3 billion, which represented 13.4% of hospital care, 4.1% of national health expenditures and 0.66% of the gross domestic product [4]. A German study involving 51 ICUs reported that total ICU costs represented 20% of overall hospital costs, 62% of which was associated with human resources [5]. In France, hospital/private clinic-related expenses represent more than 44% of all national health-care costs. Although ICU beds represent less than 5% of hospital beds in France (mainly in public hospitals), they consume around 10% of hospital costs [6]. Despite this large cost contribution, published estimates of the daily cost of an ICU stay vary considerably due to the heterogeneity of the methods used for its assessment and the profile of considered patients [7–10]. Therefore, the main objective of the CRREA (Couˆt re´el d’une journe´e en re´animation: actual cost of one day in the ICU) study was to estimate the actual daily cost of adult ICU stays in France, whatever the patient’s profile and the time of admission, using a microcosting approach.
2. Patients and methods The present study (GSK protocol 113382) was strictly observational and was approved by the French Society of Anaesthesiology and Intensive Care (Socie´te´ franc¸aise d’anesthe´sie et de re´animation [Sfar]) and the French Intensive Care Society (Socie´te´ de re´animation de langue franc¸aise [SRLF]). According to French law (law 88-1138 relative to Biomedical Research of December 20th 1988 modified on August 9th 2004), this noninterventional study did not require approval by an Ethics Committee nor informed, signed consent from patients. The study was declared to, and approved by, the National Committee for Data Processing and Civil Liberties (authorization # 1338734 obtained on March 20th 2009). However, under French law, all patients or their representatives were informed of the study by the ICU physician and could decline participation to the study. 2.1. The study design The present study was designed by a scientific committee composed of 3 intensivists (JYL, BG, JEB), 3 health economists (CP, FM, FC) and one statistician (IB) and followed 4 successive phases. 2.1.1. Phase 1: site selection and recruitment In order to well represent the overall pool of French ICUs, the present study required the participation of 23 ICUs: 12 university hospitals (centre hospitalier universitaire [CHU]), 10 general hospitals (centre hospitalier [CH]), and 1 private, non-profit, public service hospital (e´tablissements participant au service public hospitalier [PSPH]). The 3 ICUs (2 CHUs and 1 CH) managed by the 3 intensivists of the scientific committee participated in the study. The remaining ICUs were randomly selected from the French National Hospital database and stratified by hospital category (CHU, CH or PSPH). The head manager of each randomised ICU was contacted to ask his/her willingness to participate in the study. If and when he/she refused, a new randomly selected ICU was contacted until 20 ICU head managers accepted participation (Fig. 1). 2.1.2. Phase 2: patient enrolment In each ICU, 5 randomly selected adult patients were included within a one-month period between May 10th and October 16th 2009. In 22 units, only one patient per study day was included, thus resulting in 5 different day-waves over which the study was carried out. In one unit, all five patients were included on the
Fig. 1. Flow chart describing the inclusion process of ICUs into the study. BMI: body mass index; CH: centre hospitalier (French General Hospital); CHU: centre hospitalier universitaire (French University Hospital); ICU: intensive care unit; PSPH: e´tablissements participant au service public hospitalier (French private, non-profit, public service hospital); RRT: renal replacement therapy; SAPS II: simplified acute physiology score II; SOFA: sequential organ dysfunction score.
same day. Patients were included regardless of the day of the week. Patient inclusion criteria (according to the French definition of ICU patients [decree April 5th 2002]): adult patients ( 18 years of age) admitted to an ICU with a simplified acute physiology score (SAPS II) [11] 15 and requiring at least one major ICU medical procedure in accordance with the French hospital procedure-based pricing (corresponding to organ failure support; Table 1) could be included into the study. Non inclusion criteria: Patients with an expected life expectancy < 24 h, patients with a life-sustaining decision at the beginning of the daily assessment, patients for whom an ICU discharge was planned within a 24 hour period and patients previously enrolled in the present study did not participate.
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004
G Model
ACCPM-27; No. of Pages 7 J.-Y. Lefrant et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx Table 1 Medical acts in accordance with the French hospital procedure-based pricing. Requirement of infusion of dopamine or dobutamine or noradrenalin or adrenalin Mechanical ventilation 3 days Mechanical ventilation with FiO2 > 0.6 and/or PEP > 6 cmH2O Fluid loading > 50 mL/kg/24 h within one day Non-invasive ventilation > 12 h per day Renal replacement therapy for acute renal failure Red blood cell pack transfusion > ½ blood volume Fresh frozen and/or platelet transfusion for dessiminated intravascular coagulation Intracranial pressure monitoring Cardiopulmonary resuscitation Pericardial drainage External electric shock in emergency Intra-aortic conter-pulsion Mechanical ventilation in separated lungs Plasma exchange Electro-systolic stimulation probe insertion
2.1.3. Phase 3: data collection for a 24-hour period for each patient Activity data: in each ICU, the numbers of patients and different caregivers in the Unit were recorded on the study day. Patient data: age, gender, height, weight, co-morbidities, reason for admission to the unit, and SAPS II score at admission were collected [11]. The number of organ failures on the study day (Severity Organ Failure Assessment [SOFA] score [12]) was assessed. Time of care dedicated to each patient (indicated as hours:minutes [hh:min]): All hospital staff (regardless of speciality) taking part in patient management were asked to self-time the duration of care they carried out for the patient management of each intervention. These data were divided into direct (for instance, bedside-care), indirect (for instance, the time a doctor spends informing relatives of the patient’s status, and/or syringe preparation by nurses) and administrative (for instance, patient report preparation by secretaries) interventions. When an included patient was discharged from the Unit (death or transfer) during the study day, data were collected and extrapolated for a 24-hour period. Medications, materials, biological and radiological exams used for the included patient’s care were recorded. 2.1.4. Phase 4: economic assessment phase The prices of all resources identified were estimated from the hospital’s perspective (reference year 2009) based on the cost data provided by each centre.
(A) Staff time dedicated to patient care: for each hospital staff category, the cumulative time dedicated to patient care (both direct and indirect tasks) was multiplied by the average hourly wage for that staff category. (B) Other caregiver dependent times: certain indirect tasks are affected by the number of patients present in an ICU, and the time dedicated to these tasks is calculated in accordance with the proportion occupation of the ICU. These tasks included Night Watch Standby time (excluding intervention time), day standby time (excluding intervention time), the completion of transversal patient verifications during respite, cross cutting time watch, nosocomial infection prevention meetings, etc. (C) Medications: the costs of all given treatments were estimated using prices provided by each centre and the resources used by each patient. (D) Big consumables (such as haemodialysis devices, etc.) were valued using unit prices provided by each centre and the resources used by the patient. (E) The daily cost of all small consumables (such as bandages, gauzes, Renal Replacement Therapy [RRT] materials, etc.) was
3
extrapolated from the total cost of the ICU and the number of days completed. (F) The costs of laboratory tests and other investigations were directly calculated using the biological laboratory test and procedures classification (Nomenclature des actes de biologie me´dicale [NABM]; class B for biological) and the unit cost of these biological tests in each hospital. (G) The costs of other investigations were directly calculated using the common classification of medical procedures (Classification commune des actes me´dicaux [CCAM]) and the cost of these procedures in each hospital. (H) ICU administrative expenses were divided into structural costs, amortizing expenses, general logistic expenses and medical logistics. All such data were collected by the on-site study coordinator in collaboration with the financial and administrative departments. The estimation of an average patient-day cost in the ICU was based on information collected at the site level and taking into account the annual number of patient-days both at the site and ICU levels. Using these previous definitions, the actual daily cost could be separated into human resource allocation (A + B), patient-dependent expenses (A + C + D + E + F + G) and administrative expenses (B + H). 2.2. Statistical analysis Quantitative variables are expressed as mean values with standard deviations (SD) and median values with inter-quartiles [Q1–Q3] according to the distribution of the studied parameter. In order to better express studied parameters, the ranges (lowest value-highest value) were added. Qualitative variables are expressed as frequencies with percentages. The total daily cost was expressed with its 95% confidence interval. When adequate, a Pearson’s correlation coefficient was calculated. All analyses were performed using SAS system software version 9.2 (SAS Institute Inc., Cary, NC, USA) under WindowsTM. A 2-sided type 1 error rate of 0.05 was considered as the threshold for statistical significance. 3. Results 3.1. Intensive Care Units During the study period, 23 ICUs initially accepted to participate in the study, but one CHU did not include any patients. Therefore, the present study included 109 patients (one CH included 4 patients) (Fig. 1). In one CHU, the medical staff time was not assessed and the associated 5 patients were excluded for per protocol analysis, thus resulting in 104 patients in 21 ICUs (10 CHUs, 10 CHs and 1 PSPH) (Fig. 1). Measures of hospital activity in 2008 are shown in Table 2.
Table 2 Measures of activity (2008) for the studied ICUs (median values with range). Number of authorized ICU beds (n = 21) Number of opened ICU beds (n = 19) Number of admitted patients (n = 21) Number of ICU days (n = 21) Bed occupancy (%) (n = 19) Mean length of stay in ICU (days) (n = 21) Patients requiring mechanical ventilation (%)(n = 17) Patients requiring mechanical ventilation > 48 h (%)(n = 20) Mortality rate (%)(n = 18)
10 10 425 3478 100 7.2 65
(6–20) (8–18) (131–988) (1830–6900) (81–100) (5.1–20.6) (44–96)
43 (17–96) 21 (12–35)
For each studied parameter, the number of valid responses is given in parentheses.
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004
G Model
ACCPM-27; No. of Pages 7 J.-Y. Lefrant et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
4
Table 4 Patient severity.
3.2. Patients Patients (67 males, 64%) characteristics and severity are shown in Tables 3 and 4. Mean age was 62.3 14.9 years. At admission, the mean SAPS II score was 51.5 16.1 and the mean body mass index was 28.3 11.0. The median time elapsed from admission to the study day was 5.8 days [1.8–19.5]. On the study day, the mean SOFA score was 6.9 4.3. Ninety patients (87%) were mechanically ventilated, 55 (53%) were sedated and 17 (16%) required renal replacement therapy. There were 4 patients with withholding (n = 2) and withdrawing (n = 2) decisions. Two patients died without lifesustaining decisions, whereas 3 patients were prematurely discharged from the ICU. Patient characteristics were similar when the analysis included 22 intentions to treat ICUs (109 patients) (data not shown). 3.3. Staff time Over 24 hours, 29 to 186 interventions per patient were performed by different caregivers leading to a mean total time Table 3 Patient characteristics. Age (years) 62.3 14.9 65 [53–73] Sex ratio (F/M) 38/66 2 Body Mass Index (kg/m ) (n = 94) 28.3 11.0 26.5 [23–30.8] Medical history Patients with at least one 87 clinically relevant medical history 2.6 1.92 [1–4] Number of following clinically relevant medical histories 48 Hypertensiona Risk of multiresistant bacteria 38 b carriage 29 Sedentarity Diabetes mellitus 21 Tobacco abusec 21 Obliterant arteriopathy 16 Dyslipidemia 16 Coronary artery disease 15 d Chronic renal failure 15 e Alcohol abuse 15 New York Heart Association 10 (NYHA) > III 10 Chronic respiratory deficiencyf Stroke (ischemic or hemorrhagic) 5 Gastro-duodenal ulcer 5 Clinical hepatopathy 4 Main diagnosis Severe sepsis and/or septic shock 28 Acute respiratory failure 22 Haemorrhagic shock 13 Other shocks 7 Coma 6 Immediate postoperative surveillance 5 Acute exacerbation of Chronic 5 Obstructive Pulmonary Disease 5 Voluntary medicine intoxication Status epilepticus 3 Acute pancreatitis 2 Stroke 2 Miscellaneous 6
(18–87) (37%/63%) (14–99.1) 84% (0–8) 46% 37% 28% 20% 20% 15% 15% 14% 14% 14% 10% 10% 5% 5% 4% 27% 21% 13% 7% 6% 5% 5% 5% 3% 2% 2% 6%
For each studied parameter, the number of valid responses (when different from 104) is given in parentheses. For continuous variables, the results are expressed in mean value with standard deviation, median value with [inter-quartiles] and (extreme values). For categorical variables, data are expressed in absolute values with (percentages). a Systolic arterial pressure > 140 mmHg and/or diastolic arterial pressure > 90 mmHg. b Patient who was treated with antibiotics or who was admitted to hospital in the previous 6 months. c > 20 boxes per year. d Basal creatinine > 150 mMol/l or renal replacement therapy. e Daily alcohol intake > 140 g (female) or > 210 g (male). f Requiring intermittent or continuous oxygenotherapy.
51.5 16.1 50 [40–61] SAPS II score at admission SOFA score at inclusion (n = 103) 6.9 4.3 6 [4–10] Time from ICU admission to study day 19.6 39.2 5.8 [1.8–19.5] Mechanical ventilation 90 Patients with sedative drugs (n = 103) 55 17 Patients with current renal replacement therapy (n = 103) Medical act in accordance with the French hospital procedure-based tariffs Number of medical acts per patient 2.5 1.3 2 [1–3] Dopamine or dobutamine > 8 mg/kg/min 59 or nor/epinephrine infusion 58 Mechanical ventilation > 2 days with PEP < 6 cmH2O and FiO2 < 0.6 Mechanical ventilation with PEP 43 6 cmH2O or FiO2 60% Fluid loading < 50 mL/kg/24 h within 31 one day 25 Renal replacement therapy Non-invasive ventilation 24 Cardiopulmonary resuscitation 7 2 External electric shock Red blood cell pack transfusion > ½ 2 blood volume 2 Fresh frozen and/or platelet transfusion for DICa 1 Intra-aortic conter-pulsion Intracranial pressure monitoring 1
(21–97) (0–19) (0–229) 87% 53% 17%
(1–6) 57% 56% 41% 38% 24% 22% 7% 2% 2% 2% 1% 1%
For each studied parameter, the number of valid responses (when different from 104) is given in parentheses. For continuous variables, the results are expressed in mean value with standard deviation, median value with [inter-quartiles] and (extreme values). For categorical variables, data are expressed in absolute values with (percentages). a Disseminated intravascular coagulation.
spent for patient care of 13:32 05:00 hours (range = 05:08– 31:48)(nurse time = 51%) (Fig. 2). The corresponding total cost was s350 s161 (Table 5). 3.4. Daily cost of ICU stays The total daily cost per patient was s1,425 s520 (Table 5) (95% CI = s1,323 to s1,526). The different components of cost are shown in Table 5 and in Fig. 3. The cost of ICU human resources represented 43% of total daily cost. Patient-dependent expenses (s842 s521) represented 59% of the total daily cost. There was a correlation (r = 0.977, P < 0.001) between patient-dependent expenses and the total daily cost. Patient-dependent expenses were also correlated with the daily SOFA score (r = 0.271, P = 0.006) and bedside-care time (r = 0.716, P < 0.0001), but not with the SAPS II score (r = 0.129, P = 0.191). In consequence, there was a correlation between the total daily cost and the SOFA score (r = 0.230, P = 0.019)(no correlation with admission SAPS II, r = 0.094, P = 0.341). Centre dependent expenses represented 41% of the total daily cost with no correlation with total daily cost (r = 0.103, P = 0.300).
4. Discussion In this prospective observational study including 104 patients in 21 French ICUs, micro-costing was performed in order to assess the daily cost of an ICU patient. The median value of the daily cost was s1425 (95% CI = s1323 to s1526), ranging up to s3173. The major component of this cost was ICU human resources (s607), reflecting in part the time spent at the bedside by caregivers (mean value = 13:32). The second highest component was the administrative expenses assigned to the ICU (s326). Other costs
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004
G Model
ACCPM-27; No. of Pages 7 J.-Y. Lefrant et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
Fig. 2. Number of caregiver interventions per patient-day and time (expressed in hours and minutes) spent for patient care.
related to patient-dependant activities (medications, consumables, analyses) were rarely over s130 per day. However, the use of certain supplies, like haemodialysis or hemofiltration and associated consumables could largely increase this cost component. More importantly, the total daily cost was strongly related to SOFA scores, thus suggesting a relationship with patient severity (and not the severity at patient admission) and time spent at the bedside by caregivers for patient care.
5
The strength of the present study resides in the use of a microcosting approach applied to 104 randomly chosen patients in 21 different ICUs randomly selected from a representative panel of French ICUs (taking into account the proportions of public/private, University and non-University hospitals). This methodology is extremely time consuming and difficult to carry out, and 48 ICUs declined participation, and 2 were excluded at the beginning of the study organization. However, micro-costing resulted in a complete study measuring the time spent by different caregivers for patient care, the assessment of medications, consumables, and analyses used. Furthermore, the administrative expenses associated with French ICUs were assessed. To our knowledge, though Elliott [6] proposed this type of methodology in 1997, no previous, prospective study has been performed in the ICU setting with this methodology. Our findings are concordant with those reported by Tan et al. [13] in a retrospective study involving 7 European ICUs. The authors of the latter reported a total daily cost for an ICU patient ranging from s1168 to s2025. However, laboratory, imaging and administrative costs and charges were not assessed. Moreover, the time spent by different caregivers was not measured. Halpern et al. [4] reported an increase in per day ICU costs, increasing from $2698 to $3518 (+30.4%) between 2000 and 2005. This difference could probably be explained by the different methodologies used in such studies [7–9]. Intensive care units are costly on a per-patient-day basis, even if ICU beds classically represent only about 10–15% of available hospital beds [3,4]. In the USA and Netherlands, the cost of an ICU has been reported to be close to 20% of the total hospital budget [3,4,14]. However, the daily cost estimation for an ICU varies from s855 in a multicentre German study [5] to s3221 in US ICUs [4]. Many factors may explain these differences between countries. The patient case-mix, the bed occupancy rate, the density of acute care beds, the definition, the type and the use of ICUs in each country and the staff composition are evidently a major part. However, medical practice, the availability of health-care resources, the hospital payment system (public/private-mix and insurance payment) and the relative and absolute costs between national systems could influence the ICU daily cost [7,14–16]. Cost differences between studies could also be the consequence of the different methodologies used for daily cost assessment rather than real differences in daily cost [7,10,16]. Finally, the precision of the different components of cost could be questioned [13]. Though we applied a complete micro-costing approach, some limitations should be noted. First, the present ICUs and patient samples may not be representative of ICUs and ICU patients.
Table 5 Total daily cost per patient and its different components in euros (s). Values are expressed as means SDs and in medians with [inter-quartiles] and (ranges). Parameters
Mean SD
Median, [IQ], (range)
Total daily cost Time dedicated to patient care (A) Other caregiver dependent times (B) Human resources (A) + (B) Medications (C) Consumables (D) Materials for RRT Materials for RRT/patient with RRT (n = 18) Laboratory tests (E) Investigation tests (F) Sub-total patient-dependent expenses (A) + (C) + (D) + (E) + (F) Administrative expenses assigned to ICU (G) Structural costs Amortization expenses General logistics expenses Medical logistics Sub-total centre dependent expenses (B) + (G)
1425 520 350 161 257 54 607 161 125 223 140 230 45 141 261 246 139 135 88 134 842 521
1263, [1062–1628], (684–3173) 310, [250–412], (100–1,080) 245, [224–270], (158–431) 580, [508–664], (315–1289) 57, [25–137], (0–1692) 60, [35–114], (13–1533) 0, [0–0], (0–728) 168, [50–396], (2–728) 104, [60–162], (0–778) 47, [0–86], (0–962) 668, [496–1021], (159–2729)
326 104 62 32 37 29 157 95 70 60 583 113
313, [267–355], (174–613) 64, [40–78], (14–133) 30, [19–41], (7–146) 168, [99–209], (8–400) 56, [40–68], (20–311) 554, [522–625], (384–833)
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004
G Model
ACCPM-27; No. of Pages 7 J.-Y. Lefrant et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
6
€ 400
40%
€ 350
€ 350
35%
€ 326 € 300
30%
€ 257
€ 250
25%
€ 200
20%
€ 150
€ 140
€ 100
15%
€ 139
€ 125
25%
23%
18%
10%
€ 88
€ 50
10%
5%
10%
9%
6%
€0
0%
Time dedicated Other caregiver Medications to patient care dependent time
Con sumables Laboratory tests Investigation tests
Contribution to the daily total cost
Administrative expenses assigned to ICU
Average daily cost 95% CI
Fig. 3. Distribution of the daily total cost of ICU stays per patient and average cost of each component per patient-day.
However, the scheduling of units and patients was randomised and took into account the proportion of private/public institutions and University and non-University institutions in France. Moreover, the panel of 104 patients involved in the present study was similar to patient characteristics reported by a recent French study focused on the activity of 66 ICUs in the South of France and involving 605 patients [17]. The French system of care and reimbursement could impair the extrapolation of our findings to other countries. However, our results agree with those of previous studies performed in different countries. At admission, patient severity is similar to those reported in recent European observational studies, such as the EPIC and SOAP studies [18–20]. The proportion of ventilated patients and the mortality rates in the studied ICUs are in the range of the most severe patients recorded in the recent assessment of ICU activity in the United States of America [1]. Secondly, the daily cost of an ICU patient could vary according to ICU staff composition [10]. In France, Critical Care is defined by law (decree April 5th, 2002) and pertains to patients in whom at least two different organ failures are present or could occur, in whom vital prognosis is engaged and in whom technical organ support is required. Under the same law, Intensive Care is defined by a unit managing patients with only one organ failure, whereas Intensive Surveillance manages high risk patients with no organ failure (potential risk of organ decompensation). Another law (edited on the same date) determined the human resources required in Critical Care Units; at least two nurses are required to manage 5 patients, whereas 1 aid nurse is required to manage 4 patients. In terms of reimbursement, an admitted patient with a SAPS II score 15 associated with at least one reimbursementqualifying procedure (medical acts corresponding to the artificial support of an organ failure) results in a reimbursement of s850 per day spent in the ICU. This payment is added to the French DRG tariff allocated to the patient hospital stay. This previous definition helps explain our choice of inclusion and exclusion criteria. This patient population corresponds to the most severe patients admitted in ICUs in other countries [1]. The wide range of daily costs reported in the present study could be also explained by the time elapsed from admission to the time of micro-costing. Indeed, Dasta et al. [21] demonstrated that the two first days in the ICU were the most expensive. The cost of direct patient care was closely correlated to SOFA scores and, in contrast, the total daily cost was correlated with
institution-dependent costs. This finding means that patient severity directly impacts daily costs. The most serious patients required more care time, resulting in an increase in the cost of direct patient care and the total daily cost. Tan et al reported similar findings in European ICUs [13]. The ratio patient/caregiver is classically considered as an indicator of ICU quality and has been shown to be associated with fewer nosocomial infections and better quality of care [22–26]. In conclusion, this prospective observational study using a micro-costing approach demonstrated that the total daily cost of an ICU patient ranges from s684 to s3173, and is largely influenced by direct patient care. Contributorship Jean-Yves Lefrant, Bernard Garrigues, Ce´line Pribil, Isabelle Bardoulat, Fre´de´ric Courtial, Fre´de´rique Maurel, Jean-E´tienne Bazin: conception and design, analysis and interpretation of data, drafting the article or revising it critically for important intellectual content, and final approval of the version to be published. All other members of CRREA Group: collecting data and analysis. Jean-Yves Lefrant, Bernard Garrigues and Jean-E´tienne Bazin were consulted in order to define the study design, data analysis and the initial written version of the article. Fre´de´rique Maurel, Fre´de´ric Courtial and Ce´line Pribil helped with the design, the organization of the study and data management. Isabelle Bardoulat performed data management and statistical analysis. Disclosure of interest Jean-Yves Lefrant, Bernard Garrigues, and Jean-Etienne Bazin served as paid consultants for GlaxoSmithKline. Ce´line Pribil is employed by GlaxoSmithKline. Isabelle Bardoulat, Fre´de´rique Maurel, and Fre´de´ric Courtial are employed by IMS Health. Acknowledgements The authors thank the medical and nursing staff of each unit participating in the study and the help of Carey Suehs (CHU Nıˆmes) for the writing of the article.
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004
G Model
ACCPM-27; No. of Pages 7 J.-Y. Lefrant et al. / Anaesth Crit Care Pain Med xxx (2015) xxx–xxx
Funding: This study was supported by grants from GlaxoSmithKline and was contracted by IMS Health, France. References [1] Lilly CM, Zuckerman IH, Badawi O, Riker RR. Benchmark data from more than 240,000 adults that reflect the current practice of critical care in the United States. Chest 2011;140:1232–42. [2] Constantin JM, Leone M, Jaber S, Allaouchiche B, Orban JC, Cannesson M, et al. Quelle activite´ et quels personnels soignants dans 66 unite´s de re´animation du sud de la France? Ann Fr Anesth Reanim 2010;29:512–7. [3] Halpern NA, Bettes L, Greenstein R. Federal and nationwide intensive care units and healthcare costs: 1986-1992. Crit Care Med 1994;22:2001–7. [4] Halpern NA, Pastores SM. Critical care medicine in the United States 20002005: an analysis of bed numbers, occupancy rates, payer mix, and costs. Crit Care Med 2010;38:65–71. [5] Moerer O, Plock E, Mgbor U, Schmid A, Schneider H, Wischnewsky MB, et al. A German national prevalence study on the cost of intensive care: an evaluation from 51 intensive care units. Crit Care 2007;11:R69. [6] Fourrier F. Pre´sentation de la re´animation; 2011, http://www.srlf.org. [7] Elliott D. Costing intensive care services: a review of study methods, results and limitations. Aust Crit Care 1997;10:55–63. [8] Pines JM, Fager SS, Milzman DP. A review of costing methodologies in critical care studies. J Crit Care 2002;17:181–6. [9] Clement Nee Shrive FM, Ghali WA, Donaldson C, Manns BJ. The impact of using different costing methods on the results of an economic evaluation of cardiac care: microcosting vs gross-costing approaches. Health Econ 2009;18:377–88. [10] Reis Miranda D, Jegers M. Monitoring costs in the ICU: a search for a pertinent methodology. Acta Anaesthesiol Scand 2012;56:1104–13. [11] Le Gall JR, Lemeshow S, Saulnier F. A new Simplified Acute Physiology Score (SAPS II) based on a European/North American multicenter study. JAMA 1993; 270:2957–63. [12] Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A, Bruining H, et al. The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. Intensive Care Med 1996;22:707–10.
7
[13] Tan SS, Bakker J, Hoogendoorn ME, Kapila A, Martin J, Pezzi A, et al. Direct cost analysis of intensive care unit stay in four European countries: applying a standardized costing methodology. Value Health 2012;15:81–6. [14] Oostenbrink JB, Buijs-Van der Woude T, van Agthoven M, Koopmanschap MA, Rutten FF. Unit costs of inpatient hospital days. Pharmacoeconomics 2003;21: 263–71. [15] Ridley S. Critical care outcomes. Anaesthesia 2001;56:1–3. [16] Edbrooke DL, Ridley SA, Hibbert CL, Corcoran M. Variations in expenditure between adult general intensive care units in the UK. Anaesthesia 2001;56: 208–16. [17] Leone M, Ragonnet B, Alonso S, Allaouchiche B, Constantin JM, Jaber S, et al. Variable compliance with clinical practice guidelines identified in a 1-day audit at 66 French adult intensive care units. Crit Care Med 2012;40:3189–95. [18] Vincent JL, Sakr Y, Sprung CL, Ranieri VM, Reinhart K, Gerlach H, et al. Sepsis in European intensive care units: results of the SOAP study. Crit Care Med 2006;34:344–53. [19] Vincent JL, Bihari DJ, Suter PM, Bruining HA, White J, Nicolas-Chanoin MH, et al. The prevalence of nosocomial infection in intensive care units in Europe. Results of the European Prevalence of Infection in Intensive Care (EPIC) Study. JAMA 1995;274:639–44. [20] Vincent JL, Suter P, Bihari D, Bruining H. Organization of intensive care units in Europe: lessons from the EPIC study. Intensive Care Med 1997;23:1181–4. [21] Dasta JF, McLaughlin TP, Mody SH, Piech CT. Daily cost of an intensive care unit day: the contribution of mechanical ventilation. Crit Care Med 2005;33: 1266–71. [22] Hugonnet S, Chevrolet JC, Pittet D. The effect of workload on infection risk in critically ill patients. Crit Care Med 2007;35:76–81. [23] Hugonnet S, Uckay I, Pittet D:. Staffing level: a determinant of late-onset ventilator-associated pneumonia. Crit Care 2007;11:R80. [24] de Vos M, Graafmans W, Keesman E, Westert G, van der Voort PH. Quality measurement at intensive care units: which indicators should we use? J Crit Care 2007;22:267–74. [25] Garretson S. Nurse to patient ratios in American health care. Nurs Stand 2004;19:33–7. [26] Blot SI, Serra ML, Koulenti D, Lisboa T, Deja M, Myrianthefs P, et al. Patient to nurse ratio and risk of ventilator-associated pneumonia in critically ill patients. Am J Crit Care 2011;20:e1–9.
Please cite this article in press as: Lefrant J-Y, et al. The daily cost of ICU patients: A micro-costing study in 23 French Intensive Care Units. Anaesth Crit Care Pain Med (2015), http://dx.doi.org/10.1016/j.accpm.2014.09.004