- Email: [email protected]
Financial Implications of Noninvasive Positive Pressure Ventilation (NPPV)
clinical investigations in critical care Financial Implications of Noninvasive Positive Pressure Ventilation (NPPV)* Gerard J. Criner, MD, FCCP; Diane T. Kreimer, RN, RRT; Michael Tomaselli, BA, RRT; Winnie Pierson, BS; and Dean Evans, MS
Noninvasive positive pressure ventilation (NPPV) is effective in the treatment of acute and chronic respi¬
ratory failure. However, the costs and financial bal¬ and diagnosis-related group (DRG)
ance between costs
reimbursement for patients with moderate to severe respiratory failure treated with NPPV are unknown. We examined the costs and DRG reimbursement for 27 patients receiving Medicare referred with moder¬ ately severe respiratory failure for NPPV to the venti¬ lator rehabilitation unit (VRU) at Temple University is one of four Health Care Financ¬ Hospital. This unit chronic Administration ventilator-dependent dem¬ ing onstration sites that evaluates patients for NPPV, instructs them in home NPPV use, emphasizes reha¬ bilitation, and uses strict cost accounting methods. Nineteen patients were treated with NPPV in the ICU and then referred to the VRU, and 8 patients were di¬
Patients were rectly admitted for NPPV to14thehadVRU. severe 69±9 COPD, and (mean±SE) years age, 13 had various restrictive disorders. All were hyperadmission (restrictive capneic at the time of hospital 60±15; obstructive 67±3 mm Hg, PaC02) with im¬ and limited functional status. paired lung mechanics Patients averaged 8± 15 days in the ICU, or 8 ±4.7 days on the medical floor prior to VRU transfer. The VRU length length of stay averaged 20 ± 18 days, for a total of stay of 29±21 days. After implementation of NPPV, all patients had an improvement in gas exchange while spontaneously breathing and functional status that was maintained in follow-up. At 1 and 2 years of follow-up,
positive pressure ventilation (NPPV) T^Toninvasive ^ has proven effective in the treatment of acute and
.**
chronic
respiratory failure in patients with both
re¬
strictive1"4 and obstructive5"8 lung diseases. However, a significant percentage of these patients who receive
this modality are acutely ill or have severe underlying disease. Moreover, NPPV has been described as a la*From the Division of Pulmonary and Critical Care Medicine, Department of Medicine, and Departments of Finance and Administration, Temple University Hospital, Philadelphia. Supported in part by a grant 29-P-99401/3-01 from the Health Care Financing Administration (HCFA). Manuscript received October 12, 1994; revision accepted January 9, 1995. requests: Dr. Criner, Pulmonary and Critical Care Medi¬ Reprint cine, Temple University, 931 Parkinson Pavillion, Broad it Tioga
Streets,
Philadelphia, PA 19140
74% and 63% of patients were alive, respectively. Eleven patients were admitted with DRG 475 (respi¬ ratory system diagnosis with ventilator support); how¬ ever, 16 of 27 patients were admitted across five dif¬ ferent non-475 DRG codes with reimbursement rates ranging from $2,673 to $4,215. After DRG and outlier reimbursement, a total deficit of $261,948 remained (average deficit $9,701 per patient). However, indi¬ vidual patient deficits ranged from $1,113 to $32,892. of patients treated with NPPV in¬ Eighty-two percentfinancial losses that were underrecurred substantial imbursed across all assigned DRGs, including DRG 475, the highest-weighted DRG. We conclude that patients with moderate to severe respiratory failure receiving NPPV demonstrate an improvement in func¬ tional status and gas exchange that is maintained in follow-up. In addition, patients treated with NPPV in¬ cur high costs that are currently underreimbursed by the present DRG system. Newer DRG payment scales that recognize NPPV as specific treatment should be
implemented.
(CHEST 1995; 108:475-81)
units; DRG=diagnosis-related group; HCAU=arbitrary FA= Health Care Financing Administration; MEP=maximum expired pressure; MIP=maximum inspired pres¬ NPPV=noninvasive ventilation; sure;
positive
VRU=ventilator rehabilitation unit
pressure
Key words: DRG reimbursement; health-care costs; me¬ chanical ventilation; noninvasive positive pressure ventila¬ tion; respiratory failure bor-intensive therapy which indicates that some pa¬ tients treated with this modality may incur substantial health-care costs.9 At the present time, however, the costs and diagnosis-related group (DRG) reimburse¬ ment for patients treated with NPPV in moderate to severe respiratory failure are unknown. The Health Care Financing Administration (HCFA) is currently examining the costs and quality of care of patients receiving noninvasive ventilation as part of the National Chronic Ventilator Demonstration Project.10 As one of four national demonstration sites, the venti¬ lator rehabilitation unit (VRU) at Temple University School of Medicine has treated a large number of pa¬ tients receiving Medicare in respiratory failure with NPPV over the past 3 years. Herein, we review the fiCHEST /108 / 2 / AUGUST, 1995
475
Table 1.Criteria for Noninvasive Ventilation*
or
Criteria
ventilation with a volume ventilator (PLV-102, Lifecare Inc, Boulder, Colo) or bilevel positive airway pressure (BiPAP, Respironics Inc, Monroeville, Pa). Regardless of the source of NPPV, ven¬ tilation targets were geared to achieve a reduction in dyspnea, an improvement in gas exchange (ie, decreased PaC02, increased Pa02), and an increase in minute ventilation. Mask pressures, flows, ume
Clinical
1. Severe, irreversible chronic respiratory disease 2. Symptoms of nocturnal hypoventilation, including morning head¬
ache, loss of energy, enuresis, nightmares, etc
Dyspnea at rest or increased work of breathing impairing sleep or 4. Cor pulmonale due to hypoventilation and hypoxemia unrespon¬ sive to conventional treatment Physiologic 3.
sustained rest
and volumes were monitored via an in-line pressure transducer (Validyne model MP50, range±50 cm H2O, Validyne Inc, Ventura, Calif) and pneumotachograph (Fleisch 2, Gould Instruments, Cleveland, Ohio) to titrate ventilator settings to achieve airway pressure and minute ventilation targets.
capacity <25% predicted in patients with kyphoscoliosis or neuromuscular disease, but excluding patients with COPD or central hypoventilation. Patients with COPD should have FEVi
1. Vital
<25% predicted 2. PImax <50 cm H2O in COPD 3.
lar disease
or
Assessment of NPPV Effects To characterize the patient group
receiving NPPV at study en¬ try, respiratory mechanics, gas exchange, and patient functional status were measured. Forced vital capacity (FVC), forced expira¬ tory volume in 1 s (FEVi), and maximum inspired (MIP) and ex¬ pired (MEP) mouth pressures were used to assess respiratory me¬
<25 cm H2O in neuromuscu¬
PaC02 ^45 mm Hg and pH ^7.32 persisting after appropriate
of airway obstruction and metabolic disturbances or sleep desaturation (SaO£ <88%) despite conven¬ tional oxygen therapy. *Criteria modified from Braun et al24 and adopted for noninvasive ventilation by the Health Care Financing Administration (HFCA) Chronic Ventilator Demonstration Project. treatment
4. Nocturnal
chanics. Gas exchange was evaluated by arterial blood gas analysis. The functional status of all patients was measured by a seven-point functional scale as follows: l=impaired cognition; 2=awake, alert, and oriented; 3=chairbound; 4=independent in activities of daily living; 5=ambulatory, but homebound; 6=performs non-self-care activities at home (ie, cooking, housework, etc) and; 7=performs activities outside of the home (nonhospital activities).
and DRG reimbursements for patients Medicare and treated with NPPV at our site. receiving In addition, we correlate the financial costs and DRG reimbursement with clinical outcome.
nancial
whole face were used to ensure maximum patient comfort while
simultaneously minimizing mask leaks so as not to adversely affect gas exchange. The ventilator modes chosen included portable vol¬
costs
Cost Analysis Cost data were obtained from a centralized computer-based system operated by the Hospital Finance Department. Costs were broken down into direct and indirect components. Direct costs were
Methods
defined as diagnostic tests and studies, disposable supplies, and medical personnel salaries used to evaluate and treat each individ¬ ual patient. Medical personnel (ie, respiratory therapists, nurses, physical therapists) costs per patient were derived by taking the number of patient days in the unit and multiplying them times the per diem rate of reimbursement for each full-time staff position assigned only to patients in the VRU. The number and salary level of all positions assigned to the VRU for each discipline were pre¬ viously determined via a waiver cost estimate agreed to by the HCFA and Temple University Hospital. Indirect costs included overhead to cover essential institutional services (ie, security, transportation services, laundry, nondisposable equipment, etc). In addition, costs were broken down into geographic location: ICU, VRU, and general medical floor. Direct costs in the ICU and VRU were further broken down into their components (ie, pharmacy, radiologic studies, operating room expenses, respiratory therapy
Enrollment Criteria Prior to enrollment and admission to the VRU for NPPV, patients had to have maximization of medical therapy and fulfill at least two clinical and physiologic criteria for noninvasive ventilation as described in Table 1. All patients were first treated for 48 h with maximally effective doses of inhaled bronchodilators (eg, p-agonists, anticholinergic agents), systemic and inhaled corticosteroids, and oxygen. Following maximization of medical therapy, supplemental patients who then fulfilled the clinical and physiologic criteria were admitted to the VRU. NPPV Technique A variety of face mask and ventilator modalities were used to in¬ stitute NPPV. Face masks that covered the nose, nose and mouth,
Table 2.Baseline Gas Exchange, Respiratory Mechanics, and Functional Status Gas
Pa02/FIo2 Restrictive COPD
284±87 257±7
FVC, L Restrictive COPD Restrictive COPD
exchange
PaCO£, mm Hg
pH
60±15 67±3
7.37±0.04 7.35±0.01
Respiratory mechanics
1.24±0.77 1.88±0.68* 3.1J :2.1 AU 3.5: :1.7 AU
FEVi, L
MIP, cm H20
MEP, cm H20
0.84±0.43 0.77±0.33 Functional status
35 ±14 31±12
50±24 46±9
*p<0.05. 476
Clinical
Investigations in Critical Care
q*
320
fS
300
o ^
H
280 260
O
1 8, £ PL,
65
O
U
60
H
0.4
0.2
55
0.0
50
18
24
Months After Discharge
noninvasively ventilated patients receiving Medicare. Seventy-eight and 64% of all patients receiving NPPV were alive at 1 and 2 years posthospital discharge, respectively. Figure 2. Survival of all
S3
Patient age was 69±9 years; 19 were females. Overall, patients were moderately to severely ill, with 48%
7.35
ADMIT
NPPV
D/C
F/U
Figure 1. Effects of NPPV on gas exchange in all patients; NPPV
improvement in gas exchange (dagger, produced a significant with NPPV) that was maintained during sponta¬ compared p<0.05 neous breathing at discharge (D/C) and 6 months follow-up (F/U). Asterisk denotes p<0.05 compared with admission (Admit).
personnel, nursing personnel, and nursing and respiratory therapy supplies). DRG Reimbursement The DRG reimbursement per patient, including reimbursement and outlier day payments, were tabulated. by DRG assignment Outlier days are cases involving atypical length of stay or costs. An
outlier day occurs if the patient's length of stay (excluding days not covered by Medicare) exceeds the mean length of stay for discharge in that DRG by the lesser of 23 days or 3 SDs from the mean length of stay.
Statistical Description
expressed as mean±SE except where noted. A p<0.05 statistically significant.
Data are is considered
Results
Clinical Outcome Baseline demographic data of the 27 patients dem¬ onstrated that 14 patients had COPD as the cause of respiratory failure and 13 patients had a variety of re¬ strictive diseases (kyphoscoliosis [3], obesity-hypoventilation [5], neuromuscular [4], fibrothorax [1]).
(13/27) of patients demonstrating cor pulmonale by chest radiograph, or ECG. Over¬ physical examination,8±15 all, patients spent days (n=19) in the toICUVRUor floor medical (8±5 days, n=8) prior general admission. In the VRU, all patients had a mean length of stay of 20± 18 days; mean total hospitalization length of stay for all patients was 29±21 days. Table 2 shows baseline gas exchange, respiratory mechanics, and functional status in the 27 patients re¬ ceiving Medicare at study enrollment. As illustrated, patients had moderate to severely impaired gas ex¬ with reduced Pa02/Fl02, and PaC02 ranging change from 60±15 mm Hg (restrictive) to 67±3 mm Hg (COPD) depending on their underlying respiratory disease. Similarly, both groups of patients had moder¬
ate to severe impairments in respiratory mechanics and functional status. Figure 1 shows the effects of NPPV on gas exchange while spontaneously breathing on admission prior to NPPV use, during NPPV, and while spontaneously breathing at VRU discharge and prolonged follow-up. After the implementation of NPPV, all patients had an improvement in gas exchange while spontaneously breathing at hospital discharge that was maintained in 6-month follow-up. Figure 2 show's the survival of all patients receiving Medicare treated with NPPV. At 1 and 2 years of fol¬ low-up, approximately 78% and 64% of patients were
alive, respectively.
Following VRU admission and NPPV implementaCHEST /108 / 2 / AUGUST, 1995
477
tion, the seven-point functional status score for all pa¬ at hospital discharge compared with improved admission (4.19±0.18 vs 3.17±0.28 arbitrary hospital units [AU]; p<0.05) and this improvement was main¬ tained during follow-up assessments at 3 (5.61 ±0.2 AU) and 12 (5.86±0.33 AU) months. tients
Financial Data The total cost for care of all 27 patients totaled $847,272, with $600,245 spent on care in the VRU and $247,027 spent on care delivered in the ICU and gen¬ eral medical floor, respectively. The mean (±SE) cost per patient for care during their hospital stay was $31,380 ±8. Figure 3 shows a breakdown ofdirect costs for noninvasive ventilation for care delivered in the ICU (Fig 3, top), and VRU (Fig 3, bottom). As shown, in both units, a substantial percentage of the cost of care could be attributed to nursing and respiratory therapy salaries. In addition, in the VRU where we were able to apply strict cost accounting methods, a significant percentage of costs (11%) could be attrib¬ uted to nursing (7%) and respiratory therapy (4%) Pharmacy and radiograph costs were reduced supplies. in the VRU compared with the ICU, as well as oper¬ ating room services, which reflects a decrease in patient severity of illness during their VRU stay. Table 3 shows the DRG definitions, assignments, and national average reimbursement rates for patients admitted to the VRU for NPPV. Eleven patients were admitted with DRG 475 (respiratory system diagnosis with ventilator support); however, 16 of 27 patients were admitted across five different non-475 DRG codes with reimbursements ranging from $2,673 to $4,215 per admission. Total reimbursement by DRG and outlier payment for all patients receiving noninvasive ventilation totaled $514,437 and $70,887, respectively, for a total income of $585,324. After DRG and outlier reimbursement, a deficit of $261,948 remained with an average deficit of $9,701 per patient. However, individual patient defi¬ cits ranged from $1,000 to $32,892 per patient. Eighty-two percent of patients treated with NPPV incurred substantial financial losses that were underreimbursed across all assigned DRGs, including DRG 475, the highest-weighted DRG. In fact, 47% of patients responsible for substantial financial losses
Respiratory Therapy
Pharmacy
staff 18%
12%
Nursing
Personnel 45%
DRG # 87
11%
Figure 3. Breakdown of ICU (top) and VRU (bottom) direct costs category. Pharmacy, radiology, and operating room (OR) costs by were all markedly less in the VRU. Ancillary staff costs in VRU in¬ cludes salary support for nurse coordinator, secretary, clerk, and physical therapist.
Table 3.DRG Definitions, Assignments, and National Average Reimbursement Rates DRG Definition DRG No. 34 87 88
127 135 475 478
No.
Other disorders of the nervous system Pulmonary edema and respiratory failure Chronic obstructive pulmonary disease Heart failure and shock Cardiac congenital and valvular disorders Respiratory system diagnosis with ventilator support
of Patients
(n=27)
National Reimbursement Rates
$3,547 $4,215 $3,081 $3,146 $2,673 $11,149
1 9 1 4 1 11 Clinical
Investigations in Critical Care
DRG 475, while 28% and 25% were assigned 127 and 87, respectively. Therefore, DRGs assigned lost revenue in patients receiving NPPV was across all assigned DRGs and was not secondary to overrepresentation of any one particular DRG assignment.
were
Discussion
Our data reinforce previous studies1"8 and show that NPPV may have an important beneficial effect on gas
status in patients with chronic exchange and functional In our data show that pa¬ failure. addition, respiratory tients with chronic respiratory failure secondary to se¬ vere underlying disease have significant costs associ¬ NPPV. More¬
ated with their care when treated with data show that the present DRG payment scale does not adequately reflect the costs incurred in using NPPV to treat patients with moderate to severe chronic respiratory failure, and substantial financial losses can occur when treating individual patients. It is not surprising that financial underreimbursement occurs when treating patients with moderate to severe respiratory failure with NPPV. Early after the institution of the DRG prospective payment scale, several investigators described significant underreimbursement in patients receiving Medicare who were receiving conventional mechanical ventilation (ie, tra¬ cheostomy or endotracheal tube and positive pressure ventilation).11'12 In response to these investigations, new DRG codes were designed with higher reim¬ bursements to more adequately reflect the greater costs incurred when caring for patients with tracheo¬ stomy and positive pressure ventilation.13 Many factors were recognized to be important in contributing to the with treating the ventilatorhigh costs associatedThese factors include the follow¬ dependent patient. disorder precip¬ the the of (1) underlying severity ing: the need for mechanical failure; (2) itating respiratory ventilation to be commonly delivered in the expensive environment of the ICU; (3) the significant nursing and over, our
respiratory therapy personnel costs required to provide the ventilated patient; and (4) the equip¬ daily care ofassociated with the care, monitoring, and ment costs
maintenance of the ventilator itself. Noninvasive ventilation has been recently proposed to be superior to invasive ventilation (ie, tracheostomy
and mechanical ventilation) in selected patients be¬ cause it is less morbid for the patient, easier to use, and may stabilize or improve the underlying medical con¬ dition without the costs and complications of tracheo¬ stomy and invasive ventilation.14"17 However, even though noninvasive ventilation may be less expensive than invasive ventilation, the factors that contribute to the high cost of invasive ventilation are still also present in patient groups treated with noninvasive ventilation. In our patients, almost half of whom had underlying
cor pulmonale, most were bedridden or were chairbound on admission and required significant medical or nursing therapy to initially treat their underlying condition. Although these factors are not directly for the increased costs in patients who re¬ responsible ceive NPPV, this form of therapy may indicate a group of patients who need more extensive and expensive
therapy.
However, NPPV may be a significant contributing factor to the increased costs we found in treating pa¬ tients with chronic respiratory failure. To treat patients with noninvasive ventilation, a significant optimally amount of time and effort must be expended by the primary caregiver (either nurse or respiratory thera¬ to ensure that the patient-facemask interface is pist) comfortable and secure. In addition, significant effort must be expended to ensure that the appropriate ven¬ tilator settings are implemented to enhance the likeli¬ hood of patient comfort, reduced dyspnea, and im¬ Moreover, to apply this therapy proved gas exchange.8 effectively in the home environment after hospital discharge, significant time must be spent by the staff in patient education (both face mask and noninvasive ventilation source) to minimize posthospital discharge complications and to improve compliance. In our hands, approximately 2 to 2.5 weeks were required to fully evaluate, treat, and instruct the patient in nonin¬ vasive ventilation so as to achieve our targeted goals of a minimum of 4 h of NPPV use nightly. All ofthe above factors contributed to an increased length of stay and increased utilization of nursing and respiratory therapy resources that further contributed to the increased cost of NPPV therapy. Furthermore, the above suggests that NPPV, although a less invasive modality, is at least as labor intensive, or in some respects even more labor intensive than invasive ventilation. As previously re¬ by Chevrolet and colleagues,9 approximately ported 90% of one ICU nurse's time is required to deliver noninvasive ventilation effectively. Although this may be an overstatement for most patients treated with NPPV outside of the ICU, our data would seem to support the notion that noninvasive ventilation re¬ quires a significant amount of time and effort that goes unrecognized by the present DRG reimbursement
system.
Based on our data, however, the efforts and money spent to provide NPPV appear to have been beneficial for our patient group. Although one of the weaknesses of our study is that it was not a prospective, random¬ ized, controlled trial, the physiologic, functional, and survival outcomes data of our patient group compare outcomes in comparably ill favorably with reported Most of our patients were severely ill, patient groups. had cor pulmonale, or had recent respiratory failure requiring ICU admission and mechanical ventilation. In studies examining survival in patients with COPD CHEST /108 / 2 / AUGUST, 1995
479
and respiratory failure, Asmundssen and Kilburn,18 Sluiter et al, and Moser and colleagues20 found 1-year survival to range from 30 to 44%. In this patient group, 1- and 2-year survivals of 78% and 64%, respectively, with an improvement in functional status and gas exchange were extremely favorable.21 In fact, a greater than 60% 2-year survival compares favorably even with the most heroic and expensive treatment option for severe COPD, lung transplantation.22 In addition, compared with tracheostomy and positive pressure ventilation, patient morbidity, the need for ancillary services, and the cost of care are all markedly less with noninvasive ventilation.23 Therefore, although noninvasive ventilation may increase hospital costs during the initial treatment of respiratory failure, the beneficial effects of noninvasive ventilation on gas ex¬ functional status, and survival may not only change, enhance patient outcome, but also decreases hospital costs by preventing repeated hospitalizations and ICU admissions for intubation and positive pressure venti¬ lation. The preventative effect of noninvasive ventila¬ tion on repeated hospitalization, however, is outside the scope of this study and needs to be addressed in future trials. It should be recognized that the VRU as part of the HCFA Chronic Ventilator Demonstration Project gave us a
unique opportunity to use strict cost-accounting
methods to capture the costs of inpatient therapy for noninvasive ventilation. This unit is physically sepa¬ rated from the rest of the general hospital medical floor and has it own nursing and respiratory therapy supply centers. Nurses and respiratory therapists assigned to this area only treat patients receiving mechanical ven¬ tilation. This organization enabled us to determine the costs of nondisposable and disposable supplies more exactly and respiratory and nursing personnel costs in treating patients with NPPV without contaminating the cost analysis by using personnel and supplies to treat other patient groups simultaneously. In addition, this unit enabled us to capture ancillary costs needed
to treat
hospitalized patients receiving
noninvasive
ventilation (ie, salaries of unit clerk, unit secretary, nursing coordinator, unit head nurse, security, main¬
tenance).
Another strength of our study is that we character¬ ized the patients' medical conditions and clinical out¬ comes in addition to tabulating their cost of care. This is important because it allows one to correlate the cost of care to the severity of the patient's illness and enables other investigators to compare their costs of noninvasive ventilation to a comparable patient group. Furthermore, it enables one to determine the physio¬ logic improvement and survival associated with NPPV in light of its associated costs. One of the problems that our study highlights is that the present DRG system, even when using the high¬ 480
est-weighted available DRG code for respiratory fail¬ (DRG 475), does not provide sufficient revenues to treat respiratory failure in moderate to severely im¬ paired patients with NPPV. Moreover, only lowerweighted DRGs were able to be coded under the present definitions for most of our patients, which further compounds the problem of underreimbursement. Out data suggest that the present DRG system suffers from its inability to recognize NPPV as a ther¬ apy, and that modification of DRG classifications is needed to appropriately reimburse hospitals for the cost of NPPV in hospitalized patients with respiratory failure. ure
In summary, our data substantiate prior studies that show an important beneficial effect on gas exchange, functional status, and survival in patients with longterm respiratory therapy treated with NPPV. In addi¬ tion, our study shows that patients with chronic respi¬ ratory failure who have moderate to severe underlying restrictive or obstructive ventilatory defects when treated with NPPV incur substantial costs that are underreimbursed by the present DRG payment scale. Newer DRG payment scales that recognize NPPV as a specific treatment should be considered to ade¬ quately reimburse hospitals to provide this promising and effective noninvasive treatment for hypercapnic respiratory failure. ACKNOWLEDGMENTS: We would like to the ef¬ forts of Michael Beatrice and Herbert White inacknowledge the compilation of financial data, the secretarial assistance of Darlene Macon, the il¬ lustration assistance of John Travaline, and the helpful comments of Gilbert D'Alonzo.
References 1
Kerby GR, Mayer LS, Pingleton SK. Nocturnal positive pressure
via nasal mask. Am Rev
Respir Dis 1987; 135:738-40 Bye PTP, Bruderer JW, et al. Treatment of respiratory failure during sleep in patients with neuromuscular disease. Am Rev Respir Dis 1987; 135:148-52 3 Bach JR, Alba AS. Management of chronic alveolar hypoventila¬ tion by nasal ventilation. Chest 1990; 97:52-7 4 Leger P, Jennequin J, Gerard M, et al. Home positive pressure ventilation via nasal mask for patients with neuromuscular weak¬ ness or restrictive lung or chest wall disease. Respir Care 1989; 34:73-9 5 Gay PC, Patel AM, Viggiano RW, et al. Nocturnal nasal ventila¬ tion for treatment of patients with hypercapnic respiratory failure. Mayo Clin Proc 1991; 66:695-703 6 Carrey Z, Gottfried SB, Levy RD. Ventilatory muscle support in respiratory failure with nasal positive pressure ventilation. Chest 1990; 97:150-58 7 Brochard L, Isabey D, Piquet J, et al. Reversal of acute exacer¬ bations of chronic obstructive lung disease by inspiratory assis¬ tance with a face mask. N Engl J Med 1990; 323:1523-30 8 Criner GJ, Kreimer DT, Travaline J, et al. Acute and chronic ef¬ fects of noninvasive positive pressure ventilation (NPPV) on gas exchange and functional status in chronic respiratory failure (CRF). Am Rev Respir Dis 1993; 147:A985 9 Chevrolet JC, Jolliet P, Abajo B, et al. Nasal positive pressure ventilation in patients with acute respiratory failure: difficult and time-consuming procedure for nurses. Chest 1991; 100:775-82 2 Ellis ER,
Clinical
Investigations in Critical Care
10 O'Donohue WJ. Chronic ventilator-dependent units in hospitals: attacking the front end of a long-term problem. Mayo Clin Proc
11
1992; 67:198-200
Gracey DR, Gillespie DG, Nobrega FT, et al. Financial implica¬ tions of prolonged ventilator care of medicare patients under the prospective payment system: a multicenter study. Chest 1987; 91:4224-27
payment for Douglass PS, Rosen RC, Butler PW, et al. DRG recommenda¬ and ventilator patients: complications long-term tions. Chest 1987; 91:413-17 13 Gracey DR, Nobrega FT, Waessens JM, et al. Financial impli¬ cations of prolonged ventilator care under DRGs 474 and 475.
12
Chest 1987; 91:424-27
14 Hill NS. Noninvasive ventilation: does it work, for whom, and how? Am Rev Resp Dis 1993; 147:1050-55 15 Bach JR, Alba AS, Saporito LR. Intermittent positive pressure ventilation via the mouth as an alternative to tracheostomy for 257 ventilator users. Chest 1993; 103:174-82 16 Bach JR, Saporito LR. Indications and criteria for decannulation and transition from invasive to noninvasive long-term ventilatory support. Respir Care 1994; 39:515-31
17 18
Leger P. Noninvasive positive ventilation at home. Respir Care 1994; 39:501-10 Asmundssen T, Kilburn KH. Survival of acute respiratory failure:
a study of 239 episodes. Ann Intern Med 1969; 70:471-85 19 Sluiter HJ, Blokzyl EJ, Van Dijl W, et al. Conservative and res¬
pirator treatment of acute respiratory insufficiency in patients with chronic obstructive pulmonary disease: a reappraisal. Am
Rev Respir Dis 1972; 105:932-42 20 Moser KM, Shibel EM, Beamon AJ. Acute respiratory failure in
obstructive lung disease: long-term survival after treatment in an
care unit. JAMA 1973; 225:705-07 21 Hudson LD. Survival data in patients with acute and chronic lung disease requiring mechanical ventilation. Am Rev Respir Dis 1989; 140:S19-S24 22 Lung Transplantation. Report of the ATS Workshop on Lung Transplantation. Am Rev Respir Dis 1993; 147:772-76 23 Criner GJ, Tomaselli M, Kreimer D, et al. Costs and reimburse¬ ment for prolonged ventilation in a ventilator rehabilitation unit (VRU). Am Rev Respir Dis 1994; 149A641 24 Braun NMT. Intermittent mechanical ventilation. Clin Chest Med 1988; 9:153-162
intensive
American Board of Internal Medicine 1995 Certification, Recertification and Qualifying Examinations in Critical Care Medicine Registration Period: January 1 April 1,1995 -
Examination Date: November 9, 1995
Contact: Registration Section, American Board of Internal Medicine, 3624 Market Street, Philadelphia, PA 19104. Tel: 1 (800) 441-2246 or 1 (215) 243-1500; Fax: 1 (215) 382-5515.
@
critical
care
medicine
June 7 -11,1995
review
Beverly Hills, California
course
September 13 17,1995 Atlanta, Georgia
FOR INFORMATION CALL: 1-800-343-ACCP CHEST /108 / 2 / AUGUST, 1995
481
Noninvasive positive pressure ventilation (NPPV) is effective in the treatment of acute and chronic respi¬
ratory failure. However, the costs and financial bal¬ and diagnosis-related group (DRG)
ance between costs
reimbursement for patients with moderate to severe respiratory failure treated with NPPV are unknown. We examined the costs and DRG reimbursement for 27 patients receiving Medicare referred with moder¬ ately severe respiratory failure for NPPV to the venti¬ lator rehabilitation unit (VRU) at Temple University is one of four Health Care Financ¬ Hospital. This unit chronic Administration ventilator-dependent dem¬ ing onstration sites that evaluates patients for NPPV, instructs them in home NPPV use, emphasizes reha¬ bilitation, and uses strict cost accounting methods. Nineteen patients were treated with NPPV in the ICU and then referred to the VRU, and 8 patients were di¬
Patients were rectly admitted for NPPV to14thehadVRU. severe 69±9 COPD, and (mean±SE) years age, 13 had various restrictive disorders. All were hyperadmission (restrictive capneic at the time of hospital 60±15; obstructive 67±3 mm Hg, PaC02) with im¬ and limited functional status. paired lung mechanics Patients averaged 8± 15 days in the ICU, or 8 ±4.7 days on the medical floor prior to VRU transfer. The VRU length length of stay averaged 20 ± 18 days, for a total of stay of 29±21 days. After implementation of NPPV, all patients had an improvement in gas exchange while spontaneously breathing and functional status that was maintained in follow-up. At 1 and 2 years of follow-up,
positive pressure ventilation (NPPV) T^Toninvasive ^ has proven effective in the treatment of acute and
.**
chronic
respiratory failure in patients with both
re¬
strictive1"4 and obstructive5"8 lung diseases. However, a significant percentage of these patients who receive
this modality are acutely ill or have severe underlying disease. Moreover, NPPV has been described as a la*From the Division of Pulmonary and Critical Care Medicine, Department of Medicine, and Departments of Finance and Administration, Temple University Hospital, Philadelphia. Supported in part by a grant 29-P-99401/3-01 from the Health Care Financing Administration (HCFA). Manuscript received October 12, 1994; revision accepted January 9, 1995. requests: Dr. Criner, Pulmonary and Critical Care Medi¬ Reprint cine, Temple University, 931 Parkinson Pavillion, Broad it Tioga
Streets,
Philadelphia, PA 19140
74% and 63% of patients were alive, respectively. Eleven patients were admitted with DRG 475 (respi¬ ratory system diagnosis with ventilator support); how¬ ever, 16 of 27 patients were admitted across five dif¬ ferent non-475 DRG codes with reimbursement rates ranging from $2,673 to $4,215. After DRG and outlier reimbursement, a total deficit of $261,948 remained (average deficit $9,701 per patient). However, indi¬ vidual patient deficits ranged from $1,113 to $32,892. of patients treated with NPPV in¬ Eighty-two percentfinancial losses that were underrecurred substantial imbursed across all assigned DRGs, including DRG 475, the highest-weighted DRG. We conclude that patients with moderate to severe respiratory failure receiving NPPV demonstrate an improvement in func¬ tional status and gas exchange that is maintained in follow-up. In addition, patients treated with NPPV in¬ cur high costs that are currently underreimbursed by the present DRG system. Newer DRG payment scales that recognize NPPV as specific treatment should be
implemented.
(CHEST 1995; 108:475-81)
units; DRG=diagnosis-related group; HCAU=arbitrary FA= Health Care Financing Administration; MEP=maximum expired pressure; MIP=maximum inspired pres¬ NPPV=noninvasive ventilation; sure;
positive
VRU=ventilator rehabilitation unit
pressure
Key words: DRG reimbursement; health-care costs; me¬ chanical ventilation; noninvasive positive pressure ventila¬ tion; respiratory failure bor-intensive therapy which indicates that some pa¬ tients treated with this modality may incur substantial health-care costs.9 At the present time, however, the costs and diagnosis-related group (DRG) reimburse¬ ment for patients treated with NPPV in moderate to severe respiratory failure are unknown. The Health Care Financing Administration (HCFA) is currently examining the costs and quality of care of patients receiving noninvasive ventilation as part of the National Chronic Ventilator Demonstration Project.10 As one of four national demonstration sites, the venti¬ lator rehabilitation unit (VRU) at Temple University School of Medicine has treated a large number of pa¬ tients receiving Medicare in respiratory failure with NPPV over the past 3 years. Herein, we review the fiCHEST /108 / 2 / AUGUST, 1995
475
Table 1.Criteria for Noninvasive Ventilation*
or
Criteria
ventilation with a volume ventilator (PLV-102, Lifecare Inc, Boulder, Colo) or bilevel positive airway pressure (BiPAP, Respironics Inc, Monroeville, Pa). Regardless of the source of NPPV, ven¬ tilation targets were geared to achieve a reduction in dyspnea, an improvement in gas exchange (ie, decreased PaC02, increased Pa02), and an increase in minute ventilation. Mask pressures, flows, ume
Clinical
1. Severe, irreversible chronic respiratory disease 2. Symptoms of nocturnal hypoventilation, including morning head¬
ache, loss of energy, enuresis, nightmares, etc
Dyspnea at rest or increased work of breathing impairing sleep or 4. Cor pulmonale due to hypoventilation and hypoxemia unrespon¬ sive to conventional treatment Physiologic 3.
sustained rest
and volumes were monitored via an in-line pressure transducer (Validyne model MP50, range±50 cm H2O, Validyne Inc, Ventura, Calif) and pneumotachograph (Fleisch 2, Gould Instruments, Cleveland, Ohio) to titrate ventilator settings to achieve airway pressure and minute ventilation targets.
capacity <25% predicted in patients with kyphoscoliosis or neuromuscular disease, but excluding patients with COPD or central hypoventilation. Patients with COPD should have FEVi
1. Vital
<25% predicted 2. PImax <50 cm H2O in COPD 3.
lar disease
or
Assessment of NPPV Effects To characterize the patient group
receiving NPPV at study en¬ try, respiratory mechanics, gas exchange, and patient functional status were measured. Forced vital capacity (FVC), forced expira¬ tory volume in 1 s (FEVi), and maximum inspired (MIP) and ex¬ pired (MEP) mouth pressures were used to assess respiratory me¬
<25 cm H2O in neuromuscu¬
PaC02 ^45 mm Hg and pH ^7.32 persisting after appropriate
of airway obstruction and metabolic disturbances or sleep desaturation (SaO£ <88%) despite conven¬ tional oxygen therapy. *Criteria modified from Braun et al24 and adopted for noninvasive ventilation by the Health Care Financing Administration (HFCA) Chronic Ventilator Demonstration Project. treatment
4. Nocturnal
chanics. Gas exchange was evaluated by arterial blood gas analysis. The functional status of all patients was measured by a seven-point functional scale as follows: l=impaired cognition; 2=awake, alert, and oriented; 3=chairbound; 4=independent in activities of daily living; 5=ambulatory, but homebound; 6=performs non-self-care activities at home (ie, cooking, housework, etc) and; 7=performs activities outside of the home (nonhospital activities).
and DRG reimbursements for patients Medicare and treated with NPPV at our site. receiving In addition, we correlate the financial costs and DRG reimbursement with clinical outcome.
nancial
whole face were used to ensure maximum patient comfort while
simultaneously minimizing mask leaks so as not to adversely affect gas exchange. The ventilator modes chosen included portable vol¬
costs
Cost Analysis Cost data were obtained from a centralized computer-based system operated by the Hospital Finance Department. Costs were broken down into direct and indirect components. Direct costs were
Methods
defined as diagnostic tests and studies, disposable supplies, and medical personnel salaries used to evaluate and treat each individ¬ ual patient. Medical personnel (ie, respiratory therapists, nurses, physical therapists) costs per patient were derived by taking the number of patient days in the unit and multiplying them times the per diem rate of reimbursement for each full-time staff position assigned only to patients in the VRU. The number and salary level of all positions assigned to the VRU for each discipline were pre¬ viously determined via a waiver cost estimate agreed to by the HCFA and Temple University Hospital. Indirect costs included overhead to cover essential institutional services (ie, security, transportation services, laundry, nondisposable equipment, etc). In addition, costs were broken down into geographic location: ICU, VRU, and general medical floor. Direct costs in the ICU and VRU were further broken down into their components (ie, pharmacy, radiologic studies, operating room expenses, respiratory therapy
Enrollment Criteria Prior to enrollment and admission to the VRU for NPPV, patients had to have maximization of medical therapy and fulfill at least two clinical and physiologic criteria for noninvasive ventilation as described in Table 1. All patients were first treated for 48 h with maximally effective doses of inhaled bronchodilators (eg, p-agonists, anticholinergic agents), systemic and inhaled corticosteroids, and oxygen. Following maximization of medical therapy, supplemental patients who then fulfilled the clinical and physiologic criteria were admitted to the VRU. NPPV Technique A variety of face mask and ventilator modalities were used to in¬ stitute NPPV. Face masks that covered the nose, nose and mouth,
Table 2.Baseline Gas Exchange, Respiratory Mechanics, and Functional Status Gas
Pa02/FIo2 Restrictive COPD
284±87 257±7
FVC, L Restrictive COPD Restrictive COPD
exchange
PaCO£, mm Hg
pH
60±15 67±3
7.37±0.04 7.35±0.01
Respiratory mechanics
1.24±0.77 1.88±0.68* 3.1J :2.1 AU 3.5: :1.7 AU
FEVi, L
MIP, cm H20
MEP, cm H20
0.84±0.43 0.77±0.33 Functional status
35 ±14 31±12
50±24 46±9
*p<0.05. 476
Clinical
Investigations in Critical Care
q*
320
fS
300
o ^
H
280 260
O
1 8, £ PL,
65
O
U
60
H
0.4
0.2
55
0.0
50
18
24
Months After Discharge
noninvasively ventilated patients receiving Medicare. Seventy-eight and 64% of all patients receiving NPPV were alive at 1 and 2 years posthospital discharge, respectively. Figure 2. Survival of all
S3
Patient age was 69±9 years; 19 were females. Overall, patients were moderately to severely ill, with 48%
7.35
ADMIT
NPPV
D/C
F/U
Figure 1. Effects of NPPV on gas exchange in all patients; NPPV
improvement in gas exchange (dagger, produced a significant with NPPV) that was maintained during sponta¬ compared p<0.05 neous breathing at discharge (D/C) and 6 months follow-up (F/U). Asterisk denotes p<0.05 compared with admission (Admit).
personnel, nursing personnel, and nursing and respiratory therapy supplies). DRG Reimbursement The DRG reimbursement per patient, including reimbursement and outlier day payments, were tabulated. by DRG assignment Outlier days are cases involving atypical length of stay or costs. An
outlier day occurs if the patient's length of stay (excluding days not covered by Medicare) exceeds the mean length of stay for discharge in that DRG by the lesser of 23 days or 3 SDs from the mean length of stay.
Statistical Description
expressed as mean±SE except where noted. A p<0.05 statistically significant.
Data are is considered
Results
Clinical Outcome Baseline demographic data of the 27 patients dem¬ onstrated that 14 patients had COPD as the cause of respiratory failure and 13 patients had a variety of re¬ strictive diseases (kyphoscoliosis [3], obesity-hypoventilation [5], neuromuscular [4], fibrothorax [1]).
(13/27) of patients demonstrating cor pulmonale by chest radiograph, or ECG. Over¬ physical examination,8±15 all, patients spent days (n=19) in the toICUVRUor floor medical (8±5 days, n=8) prior general admission. In the VRU, all patients had a mean length of stay of 20± 18 days; mean total hospitalization length of stay for all patients was 29±21 days. Table 2 shows baseline gas exchange, respiratory mechanics, and functional status in the 27 patients re¬ ceiving Medicare at study enrollment. As illustrated, patients had moderate to severely impaired gas ex¬ with reduced Pa02/Fl02, and PaC02 ranging change from 60±15 mm Hg (restrictive) to 67±3 mm Hg (COPD) depending on their underlying respiratory disease. Similarly, both groups of patients had moder¬
ate to severe impairments in respiratory mechanics and functional status. Figure 1 shows the effects of NPPV on gas exchange while spontaneously breathing on admission prior to NPPV use, during NPPV, and while spontaneously breathing at VRU discharge and prolonged follow-up. After the implementation of NPPV, all patients had an improvement in gas exchange while spontaneously breathing at hospital discharge that was maintained in 6-month follow-up. Figure 2 show's the survival of all patients receiving Medicare treated with NPPV. At 1 and 2 years of fol¬ low-up, approximately 78% and 64% of patients were
alive, respectively.
Following VRU admission and NPPV implementaCHEST /108 / 2 / AUGUST, 1995
477
tion, the seven-point functional status score for all pa¬ at hospital discharge compared with improved admission (4.19±0.18 vs 3.17±0.28 arbitrary hospital units [AU]; p<0.05) and this improvement was main¬ tained during follow-up assessments at 3 (5.61 ±0.2 AU) and 12 (5.86±0.33 AU) months. tients
Financial Data The total cost for care of all 27 patients totaled $847,272, with $600,245 spent on care in the VRU and $247,027 spent on care delivered in the ICU and gen¬ eral medical floor, respectively. The mean (±SE) cost per patient for care during their hospital stay was $31,380 ±8. Figure 3 shows a breakdown ofdirect costs for noninvasive ventilation for care delivered in the ICU (Fig 3, top), and VRU (Fig 3, bottom). As shown, in both units, a substantial percentage of the cost of care could be attributed to nursing and respiratory therapy salaries. In addition, in the VRU where we were able to apply strict cost accounting methods, a significant percentage of costs (11%) could be attrib¬ uted to nursing (7%) and respiratory therapy (4%) Pharmacy and radiograph costs were reduced supplies. in the VRU compared with the ICU, as well as oper¬ ating room services, which reflects a decrease in patient severity of illness during their VRU stay. Table 3 shows the DRG definitions, assignments, and national average reimbursement rates for patients admitted to the VRU for NPPV. Eleven patients were admitted with DRG 475 (respiratory system diagnosis with ventilator support); however, 16 of 27 patients were admitted across five different non-475 DRG codes with reimbursements ranging from $2,673 to $4,215 per admission. Total reimbursement by DRG and outlier payment for all patients receiving noninvasive ventilation totaled $514,437 and $70,887, respectively, for a total income of $585,324. After DRG and outlier reimbursement, a deficit of $261,948 remained with an average deficit of $9,701 per patient. However, individual patient defi¬ cits ranged from $1,000 to $32,892 per patient. Eighty-two percent of patients treated with NPPV incurred substantial financial losses that were underreimbursed across all assigned DRGs, including DRG 475, the highest-weighted DRG. In fact, 47% of patients responsible for substantial financial losses
Respiratory Therapy
Pharmacy
staff 18%
12%
Nursing
Personnel 45%
DRG # 87
11%
Figure 3. Breakdown of ICU (top) and VRU (bottom) direct costs category. Pharmacy, radiology, and operating room (OR) costs by were all markedly less in the VRU. Ancillary staff costs in VRU in¬ cludes salary support for nurse coordinator, secretary, clerk, and physical therapist.
Table 3.DRG Definitions, Assignments, and National Average Reimbursement Rates DRG Definition DRG No. 34 87 88
127 135 475 478
No.
Other disorders of the nervous system Pulmonary edema and respiratory failure Chronic obstructive pulmonary disease Heart failure and shock Cardiac congenital and valvular disorders Respiratory system diagnosis with ventilator support
of Patients
(n=27)
National Reimbursement Rates
$3,547 $4,215 $3,081 $3,146 $2,673 $11,149
1 9 1 4 1 11 Clinical
Investigations in Critical Care
DRG 475, while 28% and 25% were assigned 127 and 87, respectively. Therefore, DRGs assigned lost revenue in patients receiving NPPV was across all assigned DRGs and was not secondary to overrepresentation of any one particular DRG assignment.
were
Discussion
Our data reinforce previous studies1"8 and show that NPPV may have an important beneficial effect on gas
status in patients with chronic exchange and functional In our data show that pa¬ failure. addition, respiratory tients with chronic respiratory failure secondary to se¬ vere underlying disease have significant costs associ¬ NPPV. More¬
ated with their care when treated with data show that the present DRG payment scale does not adequately reflect the costs incurred in using NPPV to treat patients with moderate to severe chronic respiratory failure, and substantial financial losses can occur when treating individual patients. It is not surprising that financial underreimbursement occurs when treating patients with moderate to severe respiratory failure with NPPV. Early after the institution of the DRG prospective payment scale, several investigators described significant underreimbursement in patients receiving Medicare who were receiving conventional mechanical ventilation (ie, tra¬ cheostomy or endotracheal tube and positive pressure ventilation).11'12 In response to these investigations, new DRG codes were designed with higher reim¬ bursements to more adequately reflect the greater costs incurred when caring for patients with tracheo¬ stomy and positive pressure ventilation.13 Many factors were recognized to be important in contributing to the with treating the ventilatorhigh costs associatedThese factors include the follow¬ dependent patient. disorder precip¬ the the of (1) underlying severity ing: the need for mechanical failure; (2) itating respiratory ventilation to be commonly delivered in the expensive environment of the ICU; (3) the significant nursing and over, our
respiratory therapy personnel costs required to provide the ventilated patient; and (4) the equip¬ daily care ofassociated with the care, monitoring, and ment costs
maintenance of the ventilator itself. Noninvasive ventilation has been recently proposed to be superior to invasive ventilation (ie, tracheostomy
and mechanical ventilation) in selected patients be¬ cause it is less morbid for the patient, easier to use, and may stabilize or improve the underlying medical con¬ dition without the costs and complications of tracheo¬ stomy and invasive ventilation.14"17 However, even though noninvasive ventilation may be less expensive than invasive ventilation, the factors that contribute to the high cost of invasive ventilation are still also present in patient groups treated with noninvasive ventilation. In our patients, almost half of whom had underlying
cor pulmonale, most were bedridden or were chairbound on admission and required significant medical or nursing therapy to initially treat their underlying condition. Although these factors are not directly for the increased costs in patients who re¬ responsible ceive NPPV, this form of therapy may indicate a group of patients who need more extensive and expensive
therapy.
However, NPPV may be a significant contributing factor to the increased costs we found in treating pa¬ tients with chronic respiratory failure. To treat patients with noninvasive ventilation, a significant optimally amount of time and effort must be expended by the primary caregiver (either nurse or respiratory thera¬ to ensure that the patient-facemask interface is pist) comfortable and secure. In addition, significant effort must be expended to ensure that the appropriate ven¬ tilator settings are implemented to enhance the likeli¬ hood of patient comfort, reduced dyspnea, and im¬ Moreover, to apply this therapy proved gas exchange.8 effectively in the home environment after hospital discharge, significant time must be spent by the staff in patient education (both face mask and noninvasive ventilation source) to minimize posthospital discharge complications and to improve compliance. In our hands, approximately 2 to 2.5 weeks were required to fully evaluate, treat, and instruct the patient in nonin¬ vasive ventilation so as to achieve our targeted goals of a minimum of 4 h of NPPV use nightly. All ofthe above factors contributed to an increased length of stay and increased utilization of nursing and respiratory therapy resources that further contributed to the increased cost of NPPV therapy. Furthermore, the above suggests that NPPV, although a less invasive modality, is at least as labor intensive, or in some respects even more labor intensive than invasive ventilation. As previously re¬ by Chevrolet and colleagues,9 approximately ported 90% of one ICU nurse's time is required to deliver noninvasive ventilation effectively. Although this may be an overstatement for most patients treated with NPPV outside of the ICU, our data would seem to support the notion that noninvasive ventilation re¬ quires a significant amount of time and effort that goes unrecognized by the present DRG reimbursement
system.
Based on our data, however, the efforts and money spent to provide NPPV appear to have been beneficial for our patient group. Although one of the weaknesses of our study is that it was not a prospective, random¬ ized, controlled trial, the physiologic, functional, and survival outcomes data of our patient group compare outcomes in comparably ill favorably with reported Most of our patients were severely ill, patient groups. had cor pulmonale, or had recent respiratory failure requiring ICU admission and mechanical ventilation. In studies examining survival in patients with COPD CHEST /108 / 2 / AUGUST, 1995
479
and respiratory failure, Asmundssen and Kilburn,18 Sluiter et al, and Moser and colleagues20 found 1-year survival to range from 30 to 44%. In this patient group, 1- and 2-year survivals of 78% and 64%, respectively, with an improvement in functional status and gas exchange were extremely favorable.21 In fact, a greater than 60% 2-year survival compares favorably even with the most heroic and expensive treatment option for severe COPD, lung transplantation.22 In addition, compared with tracheostomy and positive pressure ventilation, patient morbidity, the need for ancillary services, and the cost of care are all markedly less with noninvasive ventilation.23 Therefore, although noninvasive ventilation may increase hospital costs during the initial treatment of respiratory failure, the beneficial effects of noninvasive ventilation on gas ex¬ functional status, and survival may not only change, enhance patient outcome, but also decreases hospital costs by preventing repeated hospitalizations and ICU admissions for intubation and positive pressure venti¬ lation. The preventative effect of noninvasive ventila¬ tion on repeated hospitalization, however, is outside the scope of this study and needs to be addressed in future trials. It should be recognized that the VRU as part of the HCFA Chronic Ventilator Demonstration Project gave us a
unique opportunity to use strict cost-accounting
methods to capture the costs of inpatient therapy for noninvasive ventilation. This unit is physically sepa¬ rated from the rest of the general hospital medical floor and has it own nursing and respiratory therapy supply centers. Nurses and respiratory therapists assigned to this area only treat patients receiving mechanical ven¬ tilation. This organization enabled us to determine the costs of nondisposable and disposable supplies more exactly and respiratory and nursing personnel costs in treating patients with NPPV without contaminating the cost analysis by using personnel and supplies to treat other patient groups simultaneously. In addition, this unit enabled us to capture ancillary costs needed
to treat
hospitalized patients receiving
noninvasive
ventilation (ie, salaries of unit clerk, unit secretary, nursing coordinator, unit head nurse, security, main¬
tenance).
Another strength of our study is that we character¬ ized the patients' medical conditions and clinical out¬ comes in addition to tabulating their cost of care. This is important because it allows one to correlate the cost of care to the severity of the patient's illness and enables other investigators to compare their costs of noninvasive ventilation to a comparable patient group. Furthermore, it enables one to determine the physio¬ logic improvement and survival associated with NPPV in light of its associated costs. One of the problems that our study highlights is that the present DRG system, even when using the high¬ 480
est-weighted available DRG code for respiratory fail¬ (DRG 475), does not provide sufficient revenues to treat respiratory failure in moderate to severely im¬ paired patients with NPPV. Moreover, only lowerweighted DRGs were able to be coded under the present definitions for most of our patients, which further compounds the problem of underreimbursement. Out data suggest that the present DRG system suffers from its inability to recognize NPPV as a ther¬ apy, and that modification of DRG classifications is needed to appropriately reimburse hospitals for the cost of NPPV in hospitalized patients with respiratory failure. ure
In summary, our data substantiate prior studies that show an important beneficial effect on gas exchange, functional status, and survival in patients with longterm respiratory therapy treated with NPPV. In addi¬ tion, our study shows that patients with chronic respi¬ ratory failure who have moderate to severe underlying restrictive or obstructive ventilatory defects when treated with NPPV incur substantial costs that are underreimbursed by the present DRG payment scale. Newer DRG payment scales that recognize NPPV as a specific treatment should be considered to ade¬ quately reimburse hospitals to provide this promising and effective noninvasive treatment for hypercapnic respiratory failure. ACKNOWLEDGMENTS: We would like to the ef¬ forts of Michael Beatrice and Herbert White inacknowledge the compilation of financial data, the secretarial assistance of Darlene Macon, the il¬ lustration assistance of John Travaline, and the helpful comments of Gilbert D'Alonzo.
References 1
Kerby GR, Mayer LS, Pingleton SK. Nocturnal positive pressure
via nasal mask. Am Rev
Respir Dis 1987; 135:738-40 Bye PTP, Bruderer JW, et al. Treatment of respiratory failure during sleep in patients with neuromuscular disease. Am Rev Respir Dis 1987; 135:148-52 3 Bach JR, Alba AS. Management of chronic alveolar hypoventila¬ tion by nasal ventilation. Chest 1990; 97:52-7 4 Leger P, Jennequin J, Gerard M, et al. Home positive pressure ventilation via nasal mask for patients with neuromuscular weak¬ ness or restrictive lung or chest wall disease. Respir Care 1989; 34:73-9 5 Gay PC, Patel AM, Viggiano RW, et al. Nocturnal nasal ventila¬ tion for treatment of patients with hypercapnic respiratory failure. Mayo Clin Proc 1991; 66:695-703 6 Carrey Z, Gottfried SB, Levy RD. Ventilatory muscle support in respiratory failure with nasal positive pressure ventilation. Chest 1990; 97:150-58 7 Brochard L, Isabey D, Piquet J, et al. Reversal of acute exacer¬ bations of chronic obstructive lung disease by inspiratory assis¬ tance with a face mask. N Engl J Med 1990; 323:1523-30 8 Criner GJ, Kreimer DT, Travaline J, et al. Acute and chronic ef¬ fects of noninvasive positive pressure ventilation (NPPV) on gas exchange and functional status in chronic respiratory failure (CRF). Am Rev Respir Dis 1993; 147:A985 9 Chevrolet JC, Jolliet P, Abajo B, et al. Nasal positive pressure ventilation in patients with acute respiratory failure: difficult and time-consuming procedure for nurses. Chest 1991; 100:775-82 2 Ellis ER,
Clinical
Investigations in Critical Care
10 O'Donohue WJ. Chronic ventilator-dependent units in hospitals: attacking the front end of a long-term problem. Mayo Clin Proc
11
1992; 67:198-200
Gracey DR, Gillespie DG, Nobrega FT, et al. Financial implica¬ tions of prolonged ventilator care of medicare patients under the prospective payment system: a multicenter study. Chest 1987; 91:4224-27
payment for Douglass PS, Rosen RC, Butler PW, et al. DRG recommenda¬ and ventilator patients: complications long-term tions. Chest 1987; 91:413-17 13 Gracey DR, Nobrega FT, Waessens JM, et al. Financial impli¬ cations of prolonged ventilator care under DRGs 474 and 475.
12
Chest 1987; 91:424-27
14 Hill NS. Noninvasive ventilation: does it work, for whom, and how? Am Rev Resp Dis 1993; 147:1050-55 15 Bach JR, Alba AS, Saporito LR. Intermittent positive pressure ventilation via the mouth as an alternative to tracheostomy for 257 ventilator users. Chest 1993; 103:174-82 16 Bach JR, Saporito LR. Indications and criteria for decannulation and transition from invasive to noninvasive long-term ventilatory support. Respir Care 1994; 39:515-31
17 18
Leger P. Noninvasive positive ventilation at home. Respir Care 1994; 39:501-10 Asmundssen T, Kilburn KH. Survival of acute respiratory failure:
a study of 239 episodes. Ann Intern Med 1969; 70:471-85 19 Sluiter HJ, Blokzyl EJ, Van Dijl W, et al. Conservative and res¬
pirator treatment of acute respiratory insufficiency in patients with chronic obstructive pulmonary disease: a reappraisal. Am
Rev Respir Dis 1972; 105:932-42 20 Moser KM, Shibel EM, Beamon AJ. Acute respiratory failure in
obstructive lung disease: long-term survival after treatment in an
care unit. JAMA 1973; 225:705-07 21 Hudson LD. Survival data in patients with acute and chronic lung disease requiring mechanical ventilation. Am Rev Respir Dis 1989; 140:S19-S24 22 Lung Transplantation. Report of the ATS Workshop on Lung Transplantation. Am Rev Respir Dis 1993; 147:772-76 23 Criner GJ, Tomaselli M, Kreimer D, et al. Costs and reimburse¬ ment for prolonged ventilation in a ventilator rehabilitation unit (VRU). Am Rev Respir Dis 1994; 149A641 24 Braun NMT. Intermittent mechanical ventilation. Clin Chest Med 1988; 9:153-162
intensive
American Board of Internal Medicine 1995 Certification, Recertification and Qualifying Examinations in Critical Care Medicine Registration Period: January 1 April 1,1995 -
Examination Date: November 9, 1995
Contact: Registration Section, American Board of Internal Medicine, 3624 Market Street, Philadelphia, PA 19104. Tel: 1 (800) 441-2246 or 1 (215) 243-1500; Fax: 1 (215) 382-5515.
@
critical
care
medicine
June 7 -11,1995
review
Beverly Hills, California
course
September 13 17,1995 Atlanta, Georgia
FOR INFORMATION CALL: 1-800-343-ACCP CHEST /108 / 2 / AUGUST, 1995
481