clinical investigations in critical care

Transcription

1 clinical investigations in critical care Weaning From Prolonged Mechanical Ventilation* The Experience at a Regional Weaning Center David]. Scheinhom, M.D., F.C.C.P.; Barbara M. Arlinian, R.N., Ph.D.; and Jodi L. Catlin, B.A. Study objective: The aim of this study was to describe the facility, patient population, outcome of treatment, and survival of patients transferred to a regional weaning center (RWC) after prolonged mechanical ventilation in the ICU setting. Design: Retrospective record review. Setting: Regional weaning center. Patients: Four hundred twenty-one consecutive ventilator dependent patients were transferred from ICU care for attempted weaning over a 36-month period. Meaaurement and results: Acute catastrophic surgical, traumatic, or septic illness resulted in ventilator dependency with much greater frequency than decompensated COPD. Of the 421 patients, 116 died and 287 survived with outcome known at discharge. Of the 287 who survived, 212 were freed from ventilator support. Patients who weaned were ventilatomlependent for 46.9 :!: 2.9 days before transfer to the RWC. Almost half of those weaned we re discharged to their homes. Survival at 6 months and 1 year after discharge was 44 percent and 28 percent respectively, and it was greater for those at home than for those discharged to an extended care facility (ECF). The RWC care was approximately $1,500 per patient day less costly than ICU care, and $208 per patient day less costly than noninvasive respiratory care unit care. Cond~Uions: Selected patients who become ventilator dependent for prolonged periods in the ICU may be transferred to an RWC with the expectation of successful weaning in a majority of cases. (Chest 1994; 105: ) c (BRH ). a dedicated RWC. We describe this patie nt population, the structure and function of the facility, the outcome of treatment, the level of care to which patients were returned, and their survival at 6 months and 1 year after discharge. We also gathered information on the cost of ventilator care and subsequent care for this group of patients. atastrophic illness, superimposed on chronic lung disea~e. frequently leads to support with mechanical ventilation in an ICU setting. While more than 90 percent of patients are weaned in a f ewdays, 1.2 weaning attempts in a small percentage of patients fail repeatedly for weeks to months, asituation worsened by ensuing complications of prolonged mechanical ventilation. Economic pressures have generated interest in transfer of these patients who are "stuck on the ventilator" to lower levels of care. Options include transfer to home or an extended care facility (ECF), where the patient is consigned to long-tenn ventilator dependency. If continued weaning efforts are conte mplated, transfer to noninvasive respiratory care units (N RCU) or to regional weaning centers (RWC) is desirable. Elpem et ap ~ have reported outcome and cost at the NRCU level, and recently Gracey et als have reported results in 61 patients treated at a demonstration unit with features more like an RWC. In a 36-month period, 421 patients were transferred for attempted weaning to Barlow Respiratory Hospital from tlw Barlow Respimtnry Hospital. Los Ange les. Manuscript n <.~ ivt>d Aprill : revision rett>ived June 4. Reprir1t requl's/.~: Dr. Sclu:inlwm, Barlau: Respiratory HosJJitnl Stadium W11y, Las Angeles BRH = Barlow Respiratory Hospital; CVDU = chronic ventilator dependent unit; ECF = extended care facility; NRCU = noninvasive respiratory care unit; PRCl.J = prolonged respiratory care unit; RWC = regional weaning center METHODS Medical re<.-urds of consecutive ventilator-dependent patients admitted to BRH for attempted weaning were reviewed. The records of 421 patients admitted between May and June were reviewed and outcomes were computed. Of these, 274 patients were admitted before June, and 147 werl' admitted thereafte r. Age and sex of the patients. history of ventilator dependenty. dumtion of ventilator d ependency both before tr.msfc r and during stabilization and attempted weaning. suecess or failure of weaning. and level of placement at discharge were recorded. In detennining the cause of ventilator dependency. we adapted the familiardassifitation ofzwillich et a! modified by Spicher and White (Table I ).~ Patients with insuffident information on dumtion of ventilation before transfer to BRH were excluded only from analysis of duration of ventilator dependency. Suc.-cess in weaning was defined as liberation from mechanical vc n tilation for at least I week. If a patient was discharged within the i days after wc aning. he was St-<>red as weaned only if follow-up phone call re<.'ords documented that he remained ventilator independe nt. Patients not t'ompletely independent. such as those dischargt'<.l with noc1umal ventilation, were <.xmnted as ventilator-dependent. Weaning from Prolonged Mechanical Ventilation (Scheinhom, Artlnisn, Catlin)

2 Table 1 - ~.. «~ d i ntg o ~ ~ 1. Chronic Lung Disease Chronic obstructive pulmonary disease Other chronic lung diseases 2. Acute Lung rn-se Pneumonia Adult respiratory distress syndrome Pulmonary embolus Aspiration pneumonia Status asthmaticus Obstructing carcinoma 3. Postoperative Cardiac surgery Other thoracic Abdominal Trauma Burns Miscellaneous 4. ~ D i s e a s e Cardiopulmonary resuscitation Congestive heart failure Myocardial infarction 5. Neurologic Disease Cerebrovascular accident Guillain Bam syndrome EncepiWitis Status epilepticus 6. Other Bacterial sepsis Drug overdose Poisoning Metabolic acidosis Adapted from Zwillich et aje and Spicher and White. Separate social service/discharge planning records were reviewed to obtain survival data for all patients at 6 months and 1 year after discharge. OUTCOME PLACEMENT Prior time (days) BRH time (days) Age (yean) Sex Male Female ~ Ventilator-dependent or died (n=219) (n=202) 46.9± ± ± ± ± ± Cost of care at BRH was generated by multiplying billed charges for each ventilator-dependent day, exclusive of physician charges, by c o -t s t ~ h aratios. r g Days e of ventilator dependency and nursing unit location data were drawn from the medical record, while itemized charges were abstracted from the BRH business office charge records. Daily cost of ICU care before transfer to BRH and cost of or charges for subsequent lower levels of care in the Los Angeles area were estimated by using available published literature and by telephone survey. Data are expressed as mean ± SEM. RESULTS Age, sex, and time spent ventilator-dependent before transfer to BRH and at BRH are shown in Table 2, with outcome of weaning attempts. Our population was elderly, with a preponderance of women. Fifty-one records contained inadequate information on the duration of ventilator dependency before transfer to BRH. Those who were weaned spent a shorter time as ventilator-dependent before transfer than those who were not weaned or who died, but the difference was not significant. In comparing the earlier group of 274 patients with the later group of 147 patients, it was found that mean time of ventilator-dependency before transfer to BRH was 10 days less in the last year of data collection than in the first 2 years. Mean time ventilator-dependent at BRH SURVIVAL SURVIVAL PLACEMENT 8 MONTHS 12 MONTHS COST, $1 DAY $28.00 $ $ $ FIGURE 1. Result of weaning attempts in 421 consecutive patients admitted ventilator-dependent to BRH. The 116 patients who died and 18 patients transferred to other acute care facilities are not included. Survival time shown starts at discharge. Cost is shown when available data allowed its calculation; otherwise, changes supplied by the facilities are shown. CHEST I 106 I 2 I FEBRUARV

3 Table 3-Etiology of Ventilator Dependency Category Number of Patients Percent Chronic lung disease Acute lung disease Postoperative Cardiac disease Neurolo~c disease Other Unable to detennine was 102 ::!: 9.1 days in patients who rt>mained ventilatordependent, compared with 39 ::!: 2.6 days in those who were weaned. Those who died had mean time as ventilator-dependent of 58 ::!: 5.3 days. The acute illnesses resulting in the need for mechanical, entilation are shown in Table 3. Decompensated chronic lung disease alone accounted for only one quarter of the patients requiring mechanical ventilation. More than half had suffered a serious direct or indirect acute pulmonary insult, postsurgical or posttraumatic event. Most of those patients also had a history of underlying lung disease. The outcome of treatment, and the level of placement at discharge, are displayed in the now diagram (Fig 1). It w;l<; not possible to determine the outcome of their treatment so the 18 patients who were discharged to other acutt> care facilities were not included in outcome analysis. One hundred sixteen patients died during the BRH hospitalization. Of the patients who survived to discharge, i4 percent were weaned. Almost half of the patie nts who were weaned were able to return home. Eight patients discharged to board and care facilities were included with those discharged to horne. Subsequent survival of all 28i patients discharged to lower levels of care is also shown in Figure l. At both 6 months and 12 months post discharge we were able to contact more than 90 pe rcent of patients or their caregive rs. Sunivai at 6 months and 1 year was considerably better in the home, whether or not patients were ventilator-dependent. We were not able to collect accurate data on hospital readmission after discharge, although we found 16 patients were readmitted to BRH ventilator-dependent after prior weaning success. A, eragt- daily cost of care while receiving mechanical ventilatory support was $980 for the 421 consecutive ventilator-dependent patients. This induded time spent in our ICU, as well as in the general care areas. Mean daily ICU costs f(x ventilator patients at other medical centers have been reported to be 83,164, with daily NRCU costs for the same patients of$1,188.~ Daily ICU costs decline with decreasing acuity of illness in those who sunri, t> the ICU stay. Costs of home-care for the debilitated patient with end-stage lung disease vary depending on the type of care provided. Haggerty et al' reported costs for a hos536 pi tal-based home-eare program for patients with severe chronic obstruetive pulmonary disease (COPD) at about $22 per day. In contrast, a Veterans Administration hospital-based home-care program for severely debilitated patients reported costs of about $7 per dayy Neither of these figures includes cost of supplies or reimbursement for caregiver time. If a patient is ventilator-dependent at home within a large health maintenance organization home-care system, the ave rage cost of care is $405 per day (EA Oppenheimer, M.D., personal communication, 1991). The ECF charges for a patient with a tracheostomy are S2i5 per day, and $600 per day if ventilator-dependent. DIS(:L"SSIO:-.; Of the 421 patients transferred to our R\VC for attempted weaning, more than 70 percent sun1ved to be discharged. Of these sunivors, 7 4percent were weaned, and almost half of those patie nts were returned to their homes. Multiple factors determined this outcome, particularly the special characteristics of BRH and the patient population studied. These factors are discussed below, with comparison to other units which attempt weaning after prolonged mechanical ventilation. Facility Regional weaning centers like BRH exist in several large metropolitan areas. Barlow RH is a 49-bed, acute care licensed, diagnosis related group exempt, not-forprofit private hospital on a 26-acre site in a Los Angeles urban park setting. The hospital's main activity is to accept ventilator-dependent patients from other hospitals in southern California and stmounding states. Aftertreatment to stabilize the patient and manage ongoing disease processes, weaning from mechanical ventilation is attempted; if it is successful, acute pulmonary rehabilitation follows. Differences from the ICU setting in the community include (1) care of ventilator-dependent patients in the general care areas, by (2) C<L'>e management teams highly experienced in management of such patients, using (3) critical paths designed for this population, with (4) medical staff who are all pulmonologistlcritical care physicians whose only practice is the full-time care of these patients, and (5) teaching pulmonary/critical care fellows from two universities in an academic format typical of a university-affiliated teaching hospital. A six-bed fully-equipped ICU is used for assessing and initially treating sicker or less stable patients on admission and for treatment of new acute problems and complications. Nurse-to-patient ratio is 1:4 to 1:5 in the general care areas and 1:1 to 1:2 in our ICU. Respiratory therapist-to-patient ratio is 1:7 in the general care areas and 1:3 to 1:4 in our ICU. The dedication to, and expertise in the delivery of specialized care is renected in the mean employment duration of full-time registered nurses at BRH which is 8 years; for respiratory therapist it is 6 Weaning from Prolonged Mechanical Ventilation (Scheinhom, Artinian. Catlin)

4 years. Occupational therapy, physical therapy, and speech and swallowing therapy, all attuned to the specific problems and needs of the ventilator patient, serve both those patients who are still ventilator-dependent and those who have progressed to ambulatory status. Consultative services and surgical services if needed are provided by specialists based at nearby hospitals. Patients The ventilator-dependent patients transferred to our RWC are usually elderly, more are women than men, and most have been ventilator-dependent for more than 6 weeks. Our data show that 55 percent became ventilator-dependent after severe acute lung injury resulting from infections, postsurgical complications, or traumatic injury superimposed on chronic lung disease. Only 25 percent simply had exacerbation of COPD. Comorbid conditions related to smoking, such as coronary artery disease and peripheral vascular disease frequently initiated the process, leading to surgery in these patients. During prolonged ventilator dependency before transfer, complications of mechanical ventilation occurred which also contributed to difficulty in weaning. These "stuck on the ventilator" patients comprise only a small percentage of patients in the practice of most pulmonologists and intensivists. After weeks to months of being ventilator-dependent, the frustration of the patient, relatives, and caregivers leads to consideration of transfer if stabilization of acute medical and surgical problems allows. Patients generally have had a tracheostomy and must travel by paramedic ambulance, with a nurse and respiratory therapist on board. Patients are accepted for transfer to BRH after screening of telephone information gathered by a skilled admitting nurse. Medical information is reviewed by a physician, and financial information by business office professionals. It is of great importance in considering the outcome data to note that patients thought to have no chance of weaning, or no significant rehabilitative potential are seldom accepted. Examples, respectively, might be a patient with amyotrophic lateral sclerosis with prolonged ventilator dependency or a patient who is in a chronic vegetative state after a motor vehicle accident. On transfer to BRH, the patient is assigne<.\ to an attending physician who decides when the patient is ready for weaning, and chooses the weaning modality or combination of modes. No rigid weaning protocols were used during the months covered in this survey. Ventilator support was withdrawn using intermittent mandatory ventilation or pressure support ventilation (or both), until low levels of support were achieved. \Veaning was then completed in almost all cases using self-breathing trials of increasing duration. Outco111e There are seveml studies, compared in Table 4, reporting outcome after prolonged ventilator dependency. In each, weaning was attempted in a different type of unit. Comparisons of outcome between these units is made difficult by several factors, especially each unit's definition of"prolonged," and by data selection in which, for example. it is left unclear whether patients discharged to home were weaned or ventilator-dependent. Also, an objective measurement of re habilitative potential required for admission to the unit, which probably affects outcome as much as or more than time ventilator-dependent, was usually not available in retrospective analyses, including our own. Indiharw reported a H)-year e>.1)erience \vith lil patients (92 women, 79 men) ventilator-dependent for 55 days mean time before transfer to his prolonged respiratory care unit (PRCU). Our patients had a similar time of ventilator dependency before transfer to the RWC, and mean time to weaning (for those who were weaned) was remarkably similar at 39 days in both studies. The major difference in the PRCU population was that COPD was the cause of ventilator dependency in 57 percent of patients. In contrast, our patients had suffered catastrophic illnesses much more frequently than decompensated CO PD. If this difference means that our patients were generally "sicker," the fact that our success rate of 53 percent weaned and 25 percent returned home ventilator independent exceed PRCU values of 34 percent and 22 percent, respectively, may be the result of an "intensity of practice" factor. The RWC has received many more ventilator-dependent patients than Table 4-Compariaon ofumu Reporting ~ningfrom Prolonged Mechanicol Ventilotion Author Year Name of unit Number of patients Diagnosis precipitating ventilator dependency Dayst on ventilator before transfer to unit Dayst on ventilator in unit, to weaning %Weaned % Survival to discharge % Survival at 12 months after discharge Elpem' 1989 lndihar 1991 Gracey> 1992 Scheinhorn 1993 NRCU PRCU 171 COPD CVDU 61 COPD RWC 421 Medical-Surgical Medical 13* *Median total time mechanically ventilated in the subgroup of27 patients who had prolonged ventilator dependency. tmean. CHEST I 105 I 2 I FEBRUARY

5 the PRCU and has received them sooner after the acute insult in the most recent years. Indihar10 observed that his unit had a better weaning success rate each year. We too find an improved weaning success rate (improved to 63 percent) in the last group of patients we received. Our greater weaning success than the 32 percent success reported by Elpern et ap in their 95 NRCU patients may be explained by a combination of intake differences. The type of patient transferred was different, in that the NRCU excluded postsurgical patients and their patient diagnoses included more cardiovascular disease and neoplasm than ours. Only a subgroup of N RCU patients, 27 of the 95, really had prolonged mechanical ventilation, and even in that subgroup median ventilator-dependent time was only 13 (range 5 to 231) days total. Of the 27 patjents, 11 (41 percent) were weaned. If we assume that ventilator dependency dated from the onset of the acute illness, that illness was much more recent when transfer to the NRCU took place than when our patients were transferred to the RWC. The NRCU patients should not have been as difficult to wean, as they were not ''stuck" on the ventilator for very long. However, they may have been more acutely ill. This assumption must be the more powerful determining factor in outcome (ie, that patients closer to the catastrophic illness are less likely to survive to be weaned). The chronic ventilator-dependent unit (CVDU) at the Mayo Clinic recently reported success in weaning 87 percent of 61 patients ventilator-dependent amean time of 34 days at entry.5 Our diagnostic classification of patients' reason for ventilator dependency was similar to theirs, allowing meaningful <:omparisons. The major similarity is that surgery played an important role in decompensation leading to ventilator dependency in both groups of patients. The major difference is percentage of patients ventilator dependent because of COPD, with many more in the CVDU. Using interrupter technique to directly measure airflow obstruction in ventilator dependent patients, the Mayo group found underlying COPD in 46 percent of their patients. Only 25 percent of our patients had decompensated COPD as the primary cause of ventilator dependency, so it is tempting to conclude that the CVDU's very high rate of success in weaning may mean that such patients are easier to wean. If we, however. used the same technique to detect COPD instead ofdiagnosing it historically, we might find a similar high percentage of patients at BRH with airflow obstruction; further study is indicated. Survival to hospital discharge of 71 percent in our patients compares favorably with the PRCU's 60 percent and NRCU's 44 percent in their 13 days of ventilation group. Conversely. in the CVDU's 61 patients, all but 3 survived to discharge. These comparisons suffer from the lack of acuity of illness scales on admission to the respectiw units and the number and severity of complications throughout the patients' treatment. 538 Level of care after discharge was determined not only by the degree of patient debility, but also by patient and family resources and community availability of specialized care and available economic resources. These factors dictated that about half of weaned patients and two thirds of ventilator-dependent patients went to an extended care facility. Our survival data after hospital discharge (Fig 1) are believable in that at 6 and 12 months, respectively, we were able to contact almost all patients or their caregivers. An elderly and debilitated population is not a mobile one, so of those we were not able to contact by telephone, most probably had died. Even if every "unable to contact" patient, however, is assumed to have died, overall survival at 6 months is still 44 percent and at 1 year 28 percent. It should be noted that among our patients who survived 6 months, 64 percent were still alive at 1 year. An identical28 percent 1-year survival figure has been reported previously in 250 patients ventilated for at least 10 days.' This survival at 1 year is consistent with the conclusion that many of our patients had endstage disease at the onset. In the NRCU study, 1-year survival after hospital discharge was 48 percent, which militates for less advanced disease; more importantly, those investigators found no correlation between duration of mechanical ventilation and long-term survival.3 Cost Physicians are feeling heightened economic pressure to move ventilator-dependent patient<; to lower levels of care. This is reflected in our data showing that the last 147 patients transferred to BRH spent a mean time ventilator dependent before transfer that was 10 days less than the prior 274 patients. That earlier transfer to a specialized facility or unit results in cost savings is well documented. 4 w Cost savings in transferring patients to specialized care units are clearly of two types: ( 1) the savings engendered when the patient is transferred out of the ICU, and (2) the savings in weaning the patient successfully so subsequent care is much less expensive. The difference in cost per-patient-day between ICU, and RWC care approaches $1,500. Besides savings to a hospital in capital costs and important space considerations in the creation of an NRCU or CVDU, using an RWC to transfer patients from the ICU would save an additional $208 daily in the case of the NRCU; 4 comparison with the CVDU is not possible using published data.~ After weaning, the savings per patient day in being ventilator independent at lower levels of care, calculated from Figure 1, are about $377 at home and $325 in an ECF. Co~cLUSJONs A dedicated acute care hospital organized to function as an RWC is a valuable resource for patients who become ventilator-dependent and have weaning and rehaweaning from Prolonged Mechanical Ventilation (Scheinhom, Artinisn, C.tlln)

6 bilitative potential. Further, it provides another discharge planning alternative for hospitals with ventilator-dependent patients in a hospital ICU. We have shown that a high percentage of patients who are very long-term ventilator-dependent can be weaned in an RWC, then cared for at lower levels of care. The benefit in quality oflife for individuals who were weaned is obvious, although no objective scale was used to measure this. Because of this improved clinical outcome and the cost savings in both the transfer of those patients from the intensive care setting and after their subsequent weaning, the RWC serves as a valuable resource for society. ACKNOWLEDGMENTS: The authors thank Edward Crandall, PhD, MD and Meg Hassenpflug. RD, MS, for thoughtful critic:ism and advice in manuscript preparation. REFERENCES l Tomlinson JR, Miller KS. Lorch DC, Smith L, Reines HD, Sahn SA. A prospective <:omparison of IMV and T-piece weaning from mechanical ventilation. Chest 1989: 96: Krieger BP, Ershowsky PF. Becker DA. Cazeroglu HB. Evaluation of conventional criteria for predicting sue<:essful weaning from mechanical ventilatory support in elderly patients. Crit Care Med 1989; 17: Elpern EH, Larson R, Dougla.<>S P, Rosen RL, Bone RC. Long term outcomes for elderly survivors of prolonged ventilator assistance. Chest 1989; 96: Elpern EH, Silver MR. Rosen RL, Bone RC. The noninvasive respiratory eare unit: pattems of use and finandal implic-.1tions. Chest 1991: 99: Gracey DR, Viggiano RW. Nat!SSt'ns JM, Hubmayr RD. Silverstein MD, Koenig CE. Outmmes nf patients admitted to a chronic ventilator-dept udent unit in an at11te-c-..re hospital. Mayo Clin Proc 1989; 67: Zwillich CW. Pierson DJ, Creagh CE, Sutton FD. Schatz E, Petty T. Complications of assisted ventilation: a p r o~ s1 istudy v e of354 consecutive episodes. Am J Med 1974; 57: Spicher JE. White DP. Out<."Ome and fum:tion following prolonged mechanic-al ventilation. Arch Intern Med 1987; 147: Haggerty MC, Stockdale-Woodley R, Nair S. Respi-care: an innovative home care program for the patient with chronic obstructive pulmonary d i s c.: Chest l. ~ 1991: 100: Cummings JE. Hu tes SL, Weaver FM, Manheim LM, Conrad KJ, Nash K, Braun B, Adelman J. Cost-ciTediveness of Veterans Administration Hospital-Based I lome Care. Areh Intern Med 1990; 150: lndihar FJ. A 10-year rcpnrt of patients in a prolonged respiratory c-me unit. Mimi Mt'd 1991; 74: Indihar FJ. C o. ~ " t O m p aor rl-an> i sfor n d1ronic ventilator patients. Chest 1991; 96:260 CHEST I 105 I 2 I FEBR\JAR'I',