Ann Thorac Surg 1995;59:1074-1078
© 1995 The Society of Thoracic Surgeons
Clinical Pathways Can Be Based on Acuity, Not Diagnosis
J. Terrance Davis, MD,
Hugh D. Allen, MD,
Kirt Felver, BS,
H. Mary Rummell, MSN,
Jean D. Powers, PhD,
Daniel M. Cohen, MD
Department of Thoracic Surgery, Children's Hospital, Columbus, Ohio
 |
Abstract
|
|---|
The standardization of medical practice is gaining acceptance as a technique for controlling length of stay and hospital charges, while maintaining quality. Most clinical pathways address specific diagnoses or procedures, but we have developed a new approach in which pathways for cardiac care are based on acuity. All congenital cardiac surgical care rendered at Columbus Children's Hospital now falls within one of four such clinical pathways. This simplified approach is easy to use and has been well accepted. Our experience in a group of 107 consecutive patients treated in this fashion is described. The results of variance analyses, along with length of stay and charge data, are presented to demonstrate the degree to which resource utilization can be standardized in this widely variable group of patients whose problems were made cohesive by classification according to acuity level. We conclude that the resultant standardization offers considerable advantages for the managed care environment.
|
Introduction
|
|---|
See also page 1078.
The standardization of medical practice is gaining acceptance as a technique for reducing lengths of stay as well as hospital charges, while maintaining quality medical care. Approaches have ranged from using general sets of principles called practice guidelines [1–3] to using specific and detailed outlines for care called clinical pathways (CPs) [4, 5]. Both approaches contain elements of continuous quality improvement [6, 7]. Typically these standardized approaches have been applied to specific diagnoses or procedures.
In the area of cardiac surgery there is little published information about this approach [8, 9]. As in other disciplines, CP methodology has been either applied to specific diagnoses or procedures [8] or individually tailored to each patient [9]. We have previously reported on the use of a CP to manage a subset of low-risk patients undergoing repair of congenital cardiac defects; this resulted in decreased lengths of stay and hospital charges [10, 11]. Recently we developed four stratified CPs based on the level of care required, from simple to complex, and have managed all cardiac surgical patients using this new CP method. This approach significantly differs from others described in the literature because the CP is based on acuity rather than on specific lesions or procedures. The purpose of this study was to evaluate our initial experience with this approach. Lengths of stay, results of variance analyses, and hospital charge data were examined as indices of our ability to stratify and standardize care for this diagnostically and procedurally diverse patient population.
|
Material and Methods
|
|---|
Four pathways were developed based on anticipated acuity, the likelihood and duration of postoperative ventilatory support, and the expected length of stay. These are summarized in Table 1
.
Pathway A was designed for patients undergoing short-stay, non--open heart procedures not requiring postoperative ventilation. Closure of patent ductus arteriosus and uncomplicated pacemaker procedures are examples of the conditions treated in this pathway. Pathway B is used for the care of open heart procedure patients not requiring overnight ventilation (eg, a simple tetralogy of Fallot, without an outflow tract patch) or of patients having more complex closed heart procedures (eg, palliative procedures for complex lesions) for whom a 3- or 4-day hospitalization is anticipated. Pathway C guides the care of patients with moderately complex conditions. Patients with a more severe form of tetralogy of Fallot or those with a simple lesion (atrial septal defect) in the setting of a more complex condition such as Down's syndrome would have their care guided by this pathway. The procedure could be either an open or closed heart one, but overnight or 1 full day of ventilation and a 6- to 7-day hospitalization are expected. Pathway D is used to guide the care of patients with the most complex conditions. The number of days on the ventilator for any given patient in this pathway can vary. Nonetheless, the care given in the intensive care unit while the patient is artificially ventilated is standardized, with the exception of the frequency of laboratory studies, which is determined daily. Postextubation care is standardized as it is in the other pathways. A patient with a given lesion may have his or her care guided by any of the pathways, depending on the procedure chosen, the overall status of the patient, or the events that take place in the operating room. The initial pathway is selected by the operating surgeon. The pathway may be changed if the patient does better or worse than anticipated.
Each pathway consists of a detailed outline of all daily interventions to be rendered from the time of preadmission to discharge. As an example, the pathway for the operative day for patients in pathway B is given in Appendix 1. The pathways were developed by a multidisciplinary, physician-driven group that consisted of representatives from cardiac surgery, cardiology, intensive care medicine, and anesthesia, as well as cardiac care clinical nurses, nurses from the intensive care unit and the ward, house staff, respiratory and occupational therapists, and members of social services and administration. The details addressed for each day are as follows:
- Activity
- Diet
- Tests (electrocardiography and echocardiography)
- Blood studies
- Medications
- Nursing monitoring and assessment
- Pulmonary care
- Wound or other treatments
- Fluid or intravenous status
- Patient or parent teaching
- Discharge planning
Once collated by the development group, the proposed pathway guidelines were distributed to all participating groups for further input. Resultant suggestions were incorporated whenever possible. The CP guidelines were then incorporated into a user-friendly, small booklet, with each day for each CP printed on a separate page. The booklets were distributed, and multiple in-service meetings were arranged with house staff, nursing, and ancillary services before the initiation of the program. A date was selected for the launch of the program.
Variances were defined as any deviation from the CP. A single extra intervention (eg, an extra chest x-ray study) was considered a minor variance. An event triggering a number of minor variants (eg, reintubation that triggered extra blood gas determinations, extra chest films, and extra intensive care unit time) was deemed a major variant. A deviation from the CP resulting in less rather than more resource utilization was considered a positive variance. An example of a positive variance would be early extubation or the elimination of a study because of a better-than-anticipated postoperative course. Variances from the pathways were initially gathered by nursing staff at shift changes and were noted in the booklet or an accompanying sheet, which became available for analysis after discharge. Shortly after the program began, a research assistant specifically reviewed the chart and order sheets on a concurrent basis to ensure the capture of all variances. Variances were collated and reviewed. The pathways are evaluated and revised biannually.
The chart was used to verify hospital and intensive care unit lengths of stay. The hospital information services department provided information on the final hospital charges. If a pathway was changed during the patient's hospitalization, the data were assigned to the last designated pathway. Standard statistical methods were used to determine the median, mean, and standard deviation of each variable by pathway.
|
Results
|
|---|
One hundred seven consecutive surgical patients operated on during a 4-month period were included this program. Of the total number of patients, 16% (n = 17) were cared for by pathway A, 31% (n = 33) by pathway B, 24% (n = 26) by pathway C, and 29% (n = 31) by pathway D.
Table 2 details the findings from the variance analyses, and the relatively small number of variances per patient can be seen. As expected, the frequency of variations from the CP increases with increasing complexity of care. The table also shows that the small number of major variances was approximately offset by a similar number of positive variances.
Table 3 shows the lengths of stay per patient by pathway, and Table 4 shows the hospital charges in the same format. As expected, both variables were larger with increasing complexity of care, and were largely unpredictable for the patients in pathway D.
Figures 1 to 4 show the distribution of hospital charges for patients by pathway. The bimodal distribution in pathways A and B reflects the fact that these pathways were used for both open and closed heart procedures. The difference between the peaks represents the resource utilization associated with cardiopulmonary bypass. By far the most important determinant of hospital charges was the length of time on the ventilator. This directly determined the length of time in the intensive care unit and heavily influenced laboratory charges, particularly the need for blood gas determinations. Also accounting for some variability in the complex pathways was a variable length of time in the hospital before operation was performed. This was particularly true for the neonatal and infant groups.
View larger version (13K):
[in this window]
[in a new window]
|
Fig 1. . Distribution of hospital charges for patients in pathway A. The actual hospital charges are plotted on the X axis against the frequency of bills in that range (Y axis).
|
|
View larger version (16K):
[in this window]
[in a new window]
|
Fig 2. . Distribution of hospital charges for patients in pathway B. The axes are the same as those in Figure 1 . The bimodal distribution relates to the use of the pathway for patients undergoing both open and closed heart procedures.
|
|
View larger version (15K):
[in this window]
[in a new window]
|
Fig 3. . Distribution of hospital charges for patients in pathway C. The axes are the same as those in Figure 1. The bimodal distribution relates to the use of the pathway for patients undergoing both open and closed heart procedures.
|
|
View larger version (14K):
[in this window]
[in a new window]
|
Fig 4. . Distribution of hospital charges for patients in pathway D. The axes are the same as those in Figure 1. The variability of the charges relates to varying lengths of time on the ventilator and to the intensive care unit length of stay. The lowest charges were those incurred by patients who died in the operating room.
|
|
|
Comment
|
|---|
These data show that the surgical care given patients with congenital cardiac conditions can be standardized into four pathways based on acuity, despite the highly variable diagnostic and therapeutic nature of the field. The relatively tight clustering of hospital charges verifies the validity of pathways A through C, which constitute 71% of the group. The small percentages of pathway changes were inevitable, as there will always be some patients who respond differently than expected, and it is important to fit care to the patient, not the other way around. As expected, the data for pathway D were highly variable. The low charges and short lengths of stay in the patients in this pathway are due to early mortalities after complex procedures. Those at the upper end of the scale represent the inevitable outliers who require extraordinary support measures such as prolonged ventilation, renal dialysis, or extracorporeal membrane oxygenation. Despite the variability of the data for pathway D as a group, the standardization of care given on a day-to-day basis is a valuable feature of this program. For instance, a patient on pathway D may require 2 days or 2 weeks on the ventilator. The charges and lengths of stay will reflect this difference, and these 2 patients might seem to have little in common. Yet the care rendered for any given day on the ventilator is nicely standardized by the use of pathways. Furthermore, patients on pathway D represent only 29% of the group, and these median charges and lengths of stay still have predictive value from the standpoint of business and pricing decisions.
Our choice of four pathways, rather than three or five, might be questioned. We believed initially, and the data have borne out the fact, that the first three pathways would be followed relatively closely most of the time. The complex conditions of the patients in pathway D preclude a close prediction of costs and lengths of stay. We believed that further subdividing this group would not decrease the variability or enhance its usefulness. It is likely that the optimal number of acuity-based pathways might be different in a different specialty.
We believe the acuity-based pathways are more useful than the traditional diagnosis or procedurally based pathways. For example, the course in an otherwise normal 1-year-old child with tetralogy of Fallot who undergoes complete repair (probably pathway B) may be quite different from that in a 2-month-old infant with tetralogy of Fallot who also has Down's syndrome (probably pathway D).
Our approach is somewhat different from that proposed by Turley and colleagues [9]. Although the goals are quite similar, the methods differ. Their method involved seven sequential decision points that pertained to further intervention in all patients. Our approach directs patient care into one of four pathways and allows deviation within the pathways as necessary. Deviation is infrequent in the first three pathways. Although we believe our approach is quite simple and reproducible, either approach, when properly applied, can direct critical attention to resource utilization and attain the necessary goals of cost containment and reduced length of stay without sacrificing quality.
We believe this approach serves several useful functions. First, it has been helpful in the medical management of patients. Second, it has been useful in the management of resource utilization in an organized and justifiable manner. Third, we believe that the tracking of pathway distribution done in conjunction with cost analysis by pathway is a useful tool for pricing managed care contracts as we look toward the introduction of capitation or more standard global fees across large numbers of procedures.
This approach has improved the quality of our medical product. It has been particularly useful in minimizing the overutilization of services in a teaching environment. The fact that the expectations for the study and care of each patient are outlined in detail frees house officers from ordering batteries of studies to protect themselves. Parents particularly appreciate the predictability of the pathways, and their perception of the organization is enhanced when attending house officers, nurses, and other caregivers are all working from the same plan. Parents also expressed satisfaction at the earlier discharge of the patients. Variance analysis revealed no adverse effects from the use of this method, and no patients were readmitted for the management of complications that could have been prevented by a longer in-patient stay.
The standardization of resource utilization offered by this approach, in conjunction with new and improving methods in cost accounting, considerably upgrades our ability to assess the actual costs for service. The relatively predictable charges in 71% of the group give us hope that the costs will also be predictable. To the degree that the distribution of pathways is stable, and knowing the median charges for each group, the charges that would be generated by a group of sufficient size can be determined by simple arithmetic.
Although these data reflect our initial experience and represent a relatively small group dealt with over a short period, we have been developing this approach for many years, and find that the CP approach becomes easier and more accepted with time. However, long-term follow-up is important. We conclude that basing CPs on acuity rather than diagnosis enhances the ability to manage both patients and resources, and is therefore useful in the managed care environment.
|
Appendix 1. Pathway B: Operative Day
|
|---|
- Consults
- Cardiology and pediatric intensive care unit for postoperative care.
- Activity
- Bedrest.
- Diet
- NPO; clears 4 to 6 hours after extubation and when bowel sounds adequate.
- Tests
- Portable chest x-ray study stat on arrival in pediatric intensive care unit; 12-lead electrocardiography if indicated.
- Laboratory Work
- On arrival in pediatric intensive care unit: complete blood count (platelets, leukocytes); blood urea nitrogen/creatine; glucose; calcium; prothrombin time/partial thromboplastin time; magnesium; calcium phosphate. Every 8 hours: complete blood count (leukocytes); calcium; glucose. Arterial blood gas measurements after end-tidal carbon dioxide tension established.
- Medications
- Antibiotics given preoperatively intramuscularly or intravenously in preoperative unit and every 8 hours postoper-atively for 8 hours. Pain medication prn; Tylenol or ibupro-fen prn for fever.
- Monitoring and Assessment
- Nursing assessment and vital signs every hour. Continuous cardiac monitor/full disclosure, arterial line and cell volume, saturation monitor, and chest tube and urine output hourly; end-tidal carbon dioxide pressure.
- Pulmonary Care
- Extubation/O2
wean per respiratory therapy protocol; turn, cough, deep breathe, incentive spirometer appropriate for age.
- Treatments
- Phisoderm scrub preoperatively. Dressing/wound check postoperatively.
- Fluid Status
- Intravenous replacement three-fourths maintenance; urine output should be three-fourths crystalloid input. Transfusion as necessary per protocol.
- Teaching
- Reinforce intensive care unit admission teaching.
|
Acknowledgments
|
|---|
We acknowledge the invaluable assistance of Ms Mary Lou Naftzger, Ms Tracy Kulik, and Ms Kelly Kranz for data collection.
|
Footnotes
|
|---|
Presented at the Poster Session of the Thirty-first Annual Meeting of The Society of Thoracic Surgeons, Palm Springs, CA, Jan 30--Feb 1, 1995.
Address reprint requests to Dr Davis, Department of Cardiac Surgery, Columbus Children's Hospital, 700 Children's Dr, Columbus, OH 43205.
|
References
|
|---|
- Hayward RSA, Laupacis A. Initiating, conducting and maintaining guidelines development programs. Can Med Assoc J 1993;148:507–12.[Medline]
- Gilmore A. Clinical practice guidelines: weapons for patients, or shields for MDs? Can Med Assoc J 1993;148:429–31.[Medline]
- Battista RN, Hodge MJ. Clinical practice guidelines: between science and art. Can Med Assoc J 1993;148:385–8.[Medline]
- Nelson MS. Critical pathways in the emergency department. J Emerg Nursing 1993;19:110–4.[Medline]
- Crummer MB, Carter V. Critical pathways-the pivotal tool. J Cardiovasc Nursing 1993;7:30–7.[Medline]
- Kibbe DC, Bentz E, McLaughlin CP. Continuous quality improvement for continuity of care. J Fam Pract 1993;36:304–8.[Medline]
- Blumenthal D. Total quality management and physicians' clinical decisions. JAMA 1993;269:2775–8.[Medline]
- Jindani A, Aps C, Neville E, Sonmez B, Tun K, Williams BT. Postoperative cardiac surgical care: an alternative approach. Br Heart J 1993;69:59–64.[Abstract/Free Full Text]
- Turley K, Tyndall M, Roge C, et al. Critical pathway methodology: effectiveness in congenital heart surgery. Ann Thorac Surg 1994;58:57–65.[Abstract]
- Davis JT, Allen HD, Cohen DM. The fiscal impact of practice patterns in pediatric cardiac surgery. Am J Cardiol 1994;74:512–4.[Medline]
- Pastorek JS, Allen HD, Davis JT. Current outcomes of surgical closure of secundum atrial septal defect. Am J Cardiol 1994;74:75–7.[Medline]
This article has been cited by other articles:
|
|
|
|
 
K. Turley, M. Tyndall, K. M. Turley, C. Roge, M. Cooper, and H. Tarnoff
Cardiovascular-Radical Outcome Method is Effective in Complex Congenital Cardiac Lesions
Ann. Thorac. Surg.,
August 1, 1996;
62(2):
386 - 391.
[Abstract]
[Full Text]
|
|
|
Top
Abstract
Introduction
