Standardized Clinical Care Pathways for Major Thoracic Cases Reduce Hospital Costs Kenton J. Zehr, MD, Patty B. Dawson, RN, Stephen C. Yang, MD, and Richard F. Heitmiller, MD Division of Thoracic Surgery, The Johns Hopkins Hospital, Baltimore, Maryland Background. Standardized clinical care pathways have been developed for postoperative management in an attempt to contain costs in an era of rising health care costs and limited resources. The purpose of this study was to assess the effect of these pathways on length of stay, hospital charges, and outcome for major thoracic surgical procedures. Methods. All anatomic lung (segmentectomy, lobectomy, and pneumonectomy) and partial and complete esophageal resections performed from July 1991 to July 1997 were retrospectively analyzed for length of stay, hospital charges, and outcome. A prospectively developed database was used. Clinical care pathways were introduced in March 1994. Comparisons were made between the procedures performed before (group I) and after (group II) pathway implementation. Common to both pathways are early mobilization and prudent x-ray and laboratory analysis. In addition, the pathway for esophagectomies emphasizes overnight intubation with 24-hour intensive care unit care, and staged diet advancement. The discharge goal was postoperative day 10. For lung resection the emphasis is early postoperative extubation with overnight intensive care unit management. The discharge goal was postoperative day 7. Results. Group I esophagectomies (n 56) had significantly greater hospital charges compared with group II (n 96) ($21,977 $13,555 versus $17,919 $5,321; p < 0.04, in actual dollars) and ($29,097 $18,586 versus $19,260 $6,000; p < 0.001, in dollars adjusted for inflation) and greater length of stay (13.6 6.9 versus 9.5 2.8 days; p < 0.001). Group I lung resections (n 185) had a significantly greater length of stay compared with group II (n 241) (8.0 6.2 versus 6.4 3.8 days; p < 0.002); although charges trended downward ($13,113 $10,711 versus $12,404 $7,189; not significant) in actual dollars, charges were significantly less in dollars adjusted for inflation ($17,103 $13,211 versus $13,432 $8,056; p < 0.01). The most significant decreases in charges for esophagectomies were in miscellaneous charges (61% in dollars adjusted for inflation), pharmaceuticals (60%), laboratory (42%) and radiologic (39%) tests, physical therapy charges (35%), and routine charges (34%). For lung resections the greatest savings occurred for pharmaceuticals (38%), supplies (34%), miscellaneous charges (25%), and routine charges (22%). Mortality was similar (esophagectomies: I, 3.6%; II, 0%; lung resections: I, 0.5%; II, 0.8%; not significant). Conclusions. Introduction of standardized clinical pathways has resulted in a marked reduction of length of stay for all major thoracic surgical procedures. Total charges were reduced for both esophagectomies (34%) and lung resections (21%) with continued quality of outcome. (Ann Thorac Surg 1998;66:914 9) 1998 by The Society of Thoracic Surgeons Consistency in providing patient care is an effective way of improving quality. Physician-directed diagnostic and therapeutic plans have been shown to be crucial in this process [1]. Clinical care pathways have been developed for a wide variety of disease processes requiring medical and surgical intervention [2 7]. Pathway implementation has been shown to reduce hospital length of stay (LOS) and significantly reduce diagnostic testing. The ability of the caregivers and the patients and their families to have clear expectations has decreased the need for defensive medicine [8]. Whether these quality of care benefits translate to costs savings in an era of rising health care costs and limited Presented at the Forty-fourth Annual Meeting of the Southern Thoracic Surgical Association, Naples, FL, Nov 6 8, 1997. Address reprint requests to Dr Heitmiller, Division of Thoracic Surgery, The Johns Hopkins Hospital, 600 N Wolfe St, Baltimore, MD 21287. resources is less clear. We sought to assess the effect of these pathways on hospital charges, LOS, and outcome for major thoracic surgical procedures. Material and Methods All anatomic lung (segmentectomy, lobectomy, and pneumonectomy) and partial and complete esophageal resections performed from July 1991 to July 1997 were retrospectively analyzed for LOS, hospital charges, and outcome. A prospectively developed database was used. Clinical care pathways were introduced in March 1994. Comparisons were made between the procedures performed before (group I) and after (group II) pathway implementation. Charges were categorized and compared by routine charges, operating room charges, pharmaceuticals, radiologic tests, laboratory tests, supplies, 1998 by The Society of Thoracic Surgeons 0003-4975/98/$19.00 Published by Elsevier Science Inc PII S0003-4975(98)00662-6
Ann Thorac Surg ZEHR ET AL 1998;66:914 9 CLINICAL CARE PATHWAYS FOR THORACIC CASES 915 Table 1. Conversion Rate to Constant Charge Dollars Adjusted by Yearly Price Increase Variable Year 1991 1992 1993 1994 1995 1996 Year increase 8.5% 6.8% 6.3% 6.0% 4.1% 4.1% Rate change 1.085 1.068 1.063 1.060 1.041 1.041 % Increase 2.2% 1.5% 0.5% 0.2% 1.8% 0.0% Conversion 1.079 1.072 1.064 1.061 1.046 1.041 physical therapy charges, and miscellaneous charges. All data presented are presented as charges. The esophagectomy clinical pathway emphasizes overnight intubation and 24- to 36-hour intensive care unit care. Patients are on telemetry, and vital signs and hemodynamics are continuously monitored, including blood pressure, central venous pressure, pulmonary artery pressure and cardiac output in selected patients, O 2 saturation, and end-tidal CO 2. Analgesia is patient controlled by means of an epidural catheter or intravenously. On postoperative day (POD) 2, monitoring lines, neck drain, and Foley catheter are discontinued and the patients are transferred to the ward. On POD 3, enteral tube feedings are begun at 10 ml/h and advanced to 30 ml/h over 2 days. Supplemental O 2 is weaned and physical therapy is initiated. If a chest tube is used it is removed on POD 4. On POD 5 or 6 a video esophagogram is performed, and a four-stage esophageal diet is commenced. The discharge goal is POD 10. Perioperative antibiotics are continued for 24 hours. Medications throughout postoperative course include H 2 histamine blockers and analgesics; simethicone and stool softener are started on POD 6. Laboratory tests are performed on PODs 1, 2, and 4. The thoracotomy clinical pathway emphasizes operating room extubation and 12- to 24-hour intensive care unit care. Patients are on telemetry and their vital signs continuously monitored. Postoperative analgesia is patient controlled by means of an epidural catheter or intravenously. A postoperative chest film is obtained. Thereafter, one chest film per day is obtained until the chest drains are discontinued on POD 3 or 4. On POD 1 the patient is transferred to the ward. Oral intake is commenced. Supplemental O 2 is weaned. Laboratory testing is performed on POD 1, and only if necessary thereafter. Early mobilization is stressed beginning the night of the operation. Perioperative antibiotics are continued for 24 hours. Medications throughout postoperative course include H 2 histamine blockers and analgesics. Inhalation agents are used as needed. The goal for discharge is POD 6. Adjustment for constant charge dollars was made using the approved Maryland Health Services Cost Review Commission methodology (Table 1). This methodology adjusts charges annually by an approved price increase. The Health Services Cost Review Commission rate year is October 1 to September 30. The number was then converted to the fiscal year at The Johns Hopkins Hospital, July 1 to June 30. The charge increases were compounded yearly. Statistical analysis comparing hospital charges, LOS, and mortality rates was performed using the two-tailed Student s t test. Significance was accorded to p values of less than 0.05. The comparison of costs to charges has been relatively constant between fiscal years 1992 and 1997 (1992, 77%; 1993, 73%; 1994, 73%; 1995, 75%; 1996%, 77%; and 1997, 85%). For example, in 1992, charge 0.77 cost. Results Group I esophagectomies (n 56) were admitted to The Johns Hopkins Hospital between July 1991 and March 1994 and group II (n 96) between April 1994 and July 1997. Ages were similar (group I, 59.5 11.6 years versus group II, 57.4 13.5 years). Groups I (n 185) and II (n 241) lung resection corresponded to similar dates. Ages were not significantly different (group I, 61.8 11.7 years versus group II, 63.7 11.5 years). Procedures performed are categorized in Table 2. Group I esophagectomies (n 56) had significantly greater hospital charges compared with group II (n 96) ($21,977 $13,555 versus $17,919 $5,321; p 0.04, in actual dollars) and ($29,097 $18,586 versus $19,260 $6,000; p 0.001, in dollars adjusted for inflation) (Fig 1). Overall percent decrease was 18% in actual dollars and 34% in dollars adjusted for inflation. This translates to a savings of $389,568 and $944,352, respectively, since implementation of pathway management. Significant de- Table 2. Procedures Performed Esophagectomy Anatomic Lung Resection Type Group I (n 56) Group II (n 96) Type Group I (n 185) Group II (n 241) Partial 47 52 Segmentectomy 142 83 Total 2 17 Lobectomies 36 154 Other a 7 27 Pneumonectomies 7 4 a Other includes fundoplications, myotomies, fistula repair, and interpositional grafts.
916 ZEHR ET AL Ann Thorac Surg CLINICAL CARE PATHWAYS FOR THORACIC CASES 1998;66:914 9 Fig 1. Total hospital charges for esophagectomy patients, in actual dollars and constant dollars adjusted for inflation according to Table 1. Fig 3. Decreasing charges of esophagectomy patients year by year, in actual and constant dollars adjusted for inflation according to Table 1. creases in charges were realized in most categories (Fig 2); miscellaneous charges (61% decrease in dollars adjusted for inflation), pharmaceuticals (60%), laboratory (42%) and radiologic (39%) tests, physical therapy charges (35%), and routine charges (34%). Charges of supplies and operating room charges were not greatly affected. When charges for esophagectomies were analyzed by year there was a continuous decrease in charge from 1991 to 1997 ($24,806 $13,663 to $16,219 $3,242 in actual dollars; $35,224 $19,401 to $16,219 $3,242 in dollars adjusted for inflation) (Fig 3). Charges for group I lung resections (n 185) charges trended downward compared with group II (n 241) ($13,113 $10,711 versus $12,404 $7,189; not significant) in actual dollars; charges were significantly less in dollars adjusted for inflation ($17,103 $13,211 versus $13,432 $8,056; p 0.01) (Fig 4). This represents a savings of $170,628 in actual dollars and $854,029 in dollars adjusted for inflation. The overall percent decrease in charges was 5.4% in actual dollars and 21% in dollars adjusted for inflation. The greatest savings (Fig 5) occurred in pharmaceuticals (38% decrease in dollars adjusted for inflation), supplies (34%), miscellaneous charges (25%), and routine charges (22%). Similarly, operating room charges were unaffected by pathway implementation. There were minimal savings in the radiology and laboratory charges. When charges for lung resections were categorized by year there was a continuous decrease in charge from 1991 to 1997 ($14,553 $13,121 to $11,980 $4,904 in actual dollars; $20,665 $18,631 to $11,980 $4,904 in dollars adjusted for inflation) (Fig 6). Fig 2. Categorized hospital charges for esophagectomy patients in actual dollars (A) and in constant dollars adjusted for inflation (B). (lab laboratory charges; OR operating room charges; other miscellaneous charges; pharm pharmaceutical charges; PT physical therapy charges; rad radiology charges; supp supplies.) Fig 4. Total hospital charges for anatomic lung resection patients, in actual dollars and constant dollars adjusted for inflation according to Table 1.
Ann Thorac Surg ZEHR ET AL 1998;66:914 9 CLINICAL CARE PATHWAYS FOR THORACIC CASES 917 Fig 7. Decreasing length of stay in both esophagectomy and anatomic lung resection patients year by year. Fig 5. Categorized hospital charges for anatomic lung resection patients in actual dollars (A) and in constant dollars adjusted for inflation (B). (lab laboratory charges; OR operating room charges; other miscellaneous charges; pharm pharmaceutical charges; PT physical therapy charges; rad radiology charges; supp supplies.) Length of stay was a significant factor in charge reduction. Group I esophagectomies had greater LOS (13.6 6.9 versus 9.5 2.8 days; p 0.001). Preoperative days were reduced from 1.1 2.9 to 0.2 1.0 days (p 0.01) and postoperative stay decreased from 12.5 5.7 to 9.2 2.9 days (p 0.001). Because routine charges was the most costly category, this reduction represented 44% of the total savings. Group I lung resections had a significantly greater LOS compared with group II (8.0 6.2 versus 6.4 3.8 days; p 0.002). Preoperative stay was similar; group I had 0.37 1.84 days and group II 0.22 1.5 days (not significant). However, postoperative LOS was reduced from 7.6 5.6 to 6.1 3.4 days (p 0.01). As a result, 40% of the overall savings was realized in this Fig 6. Decreasing charges of anatomic lung resection patients year by year, in actual and constant dollars adjusted for inflation according to Table 1. category. When LOS was analyzed by year, there was an initial dramatic decline during the piloting period of the clinical pathways in 1992 and 1993 to a relative plateau with a marked decrease in standard deviation since pathway implementation (Fig 7). Mortality was similar (esophagectomies: I, 3.6%; II, 0%; lung resections: I, 0.5%; II, 0.8%; not significant). Comment Clinical care pathways, formally known as physiciandirected diagnostic and therapeutic plans [1], were developed to standardize postoperative patient care. The primary impetus to develop clinical pathways was to improve quality of care. However, in this day of rising hospital costs and capitation, there has also been tremendous pressure to reduce costs while maintaining quality [9]. Clinical care pathways have been embraced as a means to accomplish both goals. The earliest applications of clinical pathways to patients undergoing joint replacement and treatment for congestive heart failure realized a significant reproducible reduction in hospital charges while improving quality [2]. Further applications of this methodology have been shown to address both concerns in areas specific to cardiothoracic surgery; coronary artery bypass surgery [5], thoracoscopic procedures [3], congenital heart surgery [7], and pulmonary lobectomies [4]. Creation and maintenance of thoracic surgical pathways at The Johns Hopkins Hospital has been a dynamic process. There were two goals set for pathway development and for ongoing adjustments. The first goal was to find out what tests and interventions were necessary to treat the ideal patient undergoing routine esophageal and anatomic lung resections. Standardizing postoperative care with deletion of superfluous tests would explain the observed initial decrease in cost of care. The second goal was to identify limiting factors in postoperative care that largely dictate length of hospital stay for a given operative procedure, and to see what could be done to optimally manage these factors. This process would ex-
918 ZEHR ET AL Ann Thorac Surg CLINICAL CARE PATHWAYS FOR THORACIC CASES 1998;66:914 9 plain the observed continued decrease in costs. Our premise has been that many of the benchmarks for surgical care of the patient with esophageal or lung cancer are based on historic clinical practice guidelines that have been passed down, often unchallenged, through an apprenticeship training program. Challenging these guidelines through interval meetings and dynamic modification of the pathway has been a hallmark of our system. For lung resections, factors that limit LOS include pain control, air leaks, supplemental oxygen requirement, and respiratory care. For esophagectomies, factors that limit LOS include pain control, respiratory care, anastomotic healing, and subsequent oral alimentation. Therefore, measures to improve postoperative pain control and respiratory care, techniques to reduce lung air leaks, better insight into the process of anastomotic healing, all would be expected to reduce LOS without compromising safety of care. Our data suggest the ongoing dynamic nature of the pathway and its effect on charges. In Figures 3 and 6 the charges per year show a continuous decline both in actual dollars and dollars adjusted for inflation. Although the most rapid decreases occurred in the early 90s during the time when cost containment began to be a priority, ongoing cost reduction has occurred because of changes resulting from feedback from pathway caregivers. Pathways are particularly applicable to surgical services in large teaching hospitals where there is a steady stream of residents and fellows taking care of patients postoperatively. In these systems there is tremendous variability in what tests the caregiver believes needs to be obtained to manage the patient appropriately. Most of the interventions and tests thought necessary by caregivers are directly controlled by physicians but are categorized as laboratory, radiologic, pharmaceutical, and operating room charges. Indirectly, routine charges are under physician control, being profoundly influenced by LOS. At the New York Hospital Cornell Medical Center, 70% of the total cost in the hospital was found to be physician controlled [10]. Loop [11] emphasizes the physician as the utilization manager in an era when hospital costs are increasing at 8% annually and reimbursements have failed to increase accordingly. Our physiciandirected clinical pathway methodology addressed cost containment in all areas. The reduction in variability of charges is evidenced by the decrease in standard deviation in all categories since introduction of pathways (Figs 2, 5). Residents and fellows learn a consistent pattern of care as opposed to a trial and error approach repeated on a monthly basis as they rotate on the service. The reduced LOS was the most significant factor in charge reduction. This is manifested in the routine charge category. Forty-four percent and 40% of the total charge reduction was realized in routine charges in esophogectomies and lung resections, respectively. Several changes were responsible for these savings. Esophagectomy patients are now admitted on the same day as the operation. The bowel preparation is done as an outpatient procedure. Several days of hospital stay of the esophagectomy patients have been reduced postoperatively by acceleration of diet advancement. In the lung resection patients postoperative LOS has been reduced by earlier discontinuation of chest tubes. The preoperative days had been eliminated before clinical pathway introduction. This was partially responsible for the decreased charges realized in 1992 and 1993 seen in Figure 6. Indirectly, improved pain management and its associated improved pulmonary function and earlier selfsufficiency has readied patients for earlier discharge. When evaluated by the year, LOS had a parallel downward trend compared with charges in both categories (Fig 7). Again, the standarization of care is evident with a marked decrease in the standard deviation in LOS since 1994. There were several areas that were not as clearly affected by clinical pathway management: operating room charges, supplies, and physical therapy charges for esophagectomy patients; and operating room charges, radiologic and laboratory charges, and physical therapy charges for lung resection patients. Operating room charges, supply charges, and physical therapy charges were not under the influence of the physicians directing care of the thoracic patient. Charges occurring in these areas have been determined by personnel in charge of operating room supplies and the physical therapy division. These are areas that are currently being targeted for cost reduction in pathway meetings. With regards to radiologic and laboratory charges in the lung resection patients, significant reduction in utilization of these resources took place in the 2 years before official clinical pathway introduction. They had already nearly approached their limit of appropriate utilization. Mortality has remained unaffected by these substantial changes in clinical practice. There have been no deaths in esophagectomy patients since introduction of clinical pathways and 0.8% or 2 patient deaths in the lung resection group. This maintenance of quality has been demonstrated in other similar series managed by clinical pathways [3 5, 7]. This study presents clinical pathway methodology as a dynamic process that has been effective in producing significant cost reduction for major thoracic surgical procedures. Total charges were reduced for both esophagectomies (18% in actual dollars, 34% in dollars adjusted for inflation) and lung resections (5.4%, 21%) with continued quality of outcome. References 1. Hart RI, Musfeldt CM. MD-directed critical pathways: it s time. Hospitals 1992;66:56. 2. Musfeldt C, Hart RI. Physician-directed diagnostic and therapeutic plans: a quality cure for America s health-care crisis. J Soc Health Syst 1993;4:80 8. 3. Patton MD, Scaerf R. Thoracotomy, critical pathway, and clinical outcomes. Cancer Pract 1995;3:286 94. 4. Wright CD, Wain JC, Grillo HC, Moncure AC, Macaluso SM, Mathisen DJ. Pulmonary lobectomy patient care pathway: a model to control cost and maintain quality. Ann Thorac Surg 1997;6:299 302. 5. Velasco FT, Ko W, Rosengart T, et al. Cost containment in cardiac surgery: results with a critical pathway for coronary
Ann Thorac Surg ZEHR ET AL 1998;66:914 9 CLINICAL CARE PATHWAYS FOR THORACIC CASES 919 bypass surgery at the New York Hospital Cornell Medical Center. Best Pract Benchmarking Healthcare 1996;1:21 8. 6. Cohen J, Stock M, Anderson P, Everts E. Critical pathways for head and neck surgery. Development and implementation. Arch Otolaryngol Head Neck Surg 1997;123:11 4. 7. Turley K, Tyndall M, Roge C, et al. Critical pathway methodology: effectiveness in congenital heart surgery. Ann Thorac Surg 1994;58:57 63. 8. Shulkin DJ, Ferniany IW. The effect of developing patient compendiums for critical pathways on patient satisfaction. Am J Med Qual 1996;11:43 5. 9. Denton TA, Chaux A, Matloff JM. A cardiothoracic surgery information system for the next century: implications for managed care. Ann Thorac Surg 1995;59:486 93. 10. Skinner DB. Are cost considerations contradictory to quality? Ann Thorac Surg 1995;60:1498 9. 11. Loop FD. You are in charge of cost. Ann Thorac Surg 1995;60: 1509 12. DISCUSSION DR VICTOR F. TRASTEK (Rochester, MN): I enjoyed your discussion very much. The Johns Hopkins group has done a great job in cost reduction in the general thoracic arena. Certainly, you have shown that physician-directed involvement can reduce cost in the care of our patients and keep quality up. I have an observation: When we looked at a similar situation for lung resection, we found that 80% of our costs were in the first 24 hours of the hospital stay, ie, operating room and first night in the intensive care unit. The last days of the stay were not very costly, and once you got down to 6 days, there was not much more to be gained. You mentioned that there was little a surgeon can do about the operating room cost. I would disagree as the amount of time spent in the operating room does influence cost. Therefore, what have you done to look at your operating room costs? What opportunities do you see to reduce costs in the operating room? DR ZEHR: I certainly agree with you that there are changes the surgeon can make in the operating room to reduce costs. As the surgeon becomes more experienced with a particular procedure, the operating room time decreases. We have noted that our time spent in the operating room has dramatically decreased over the last 3 years in the esophagectomy series. Thus, we have realized some decrease in charges for these patients. However, overall, we were unable to show that our current pathway management has affected our operation room charges. Several categories of costs evaluated in our study are not under direct control of the thoracic surgeons in our hospital, specifically, operating room charges, supplies, and physical therapy charges. Thus these categories were not targeted in the initial development of our clinical pathways. To get a handle on these areas will require involvement of the operating room personnel, administrators, and such, who are responsible for ordering supplies and delegating operating room time. Currently we are targeting these areas for possible cost management in our pathway meetings. DR ALLEN J. TOGUT (Wilkes-Barre, PA): I enjoyed your presentation. You used cost and charges interchangeably. Do you know what the relationship between actual cost to the institution and charges is? Is it something like 70% or is it much greater? DR ZEHR: I apologize for the use of the words costs and charges as equivalent terms during my presentation. The data presented in this paper are all charges. The differential between charges and costs closely parallel each other at The Johns Hopkins Hospital. Costs have varied from 77% of charges in fiscal year 1992 to 84% of charges in fiscal year 1997. DR JOSEPH S. MCLAUGHLIN (Baltimore, MD): That was a wonderful paper. I really enjoyed it. The question I have is not related to the money but to the satisfaction of the patient population who get their chest tubes out early and get up early and do lots of other things that in the past they did not do. Do you have any data speaking to patients satisfaction in this era of rapid discharge and so forth? I enjoyed your paper very much. DR ZEHR: Thank you, Dr McLaughlin. We do not have these data. We can offer our anecdotal experience, which has not observed any increased patient dissatisfaction despite the decreased length of stay in the hospital. Indeed they enjoy the fact that they are going home earlier as opposed to being stuck in a hospital bed. Clinical pathways in thoracic surgery are relatively new. However, the issue of patient satisfaction with clinical pathways has been evaluated in coronary artery bypass grafting at the Hospital of the University of Pennsylvania. There was a trend toward improved patient satisfaction with clinical pathway management albeit not statistically significant. I believe that our observed patient satisfaction with clinical pathway management is a result of patients having an idea where their care is heading postoperatively. They can look at the chart on the wall outside their door and see what is supposed to happen in the next several days, as opposed to everything being a black box.