The efficacy of routine central venous monitoring in major head and neck surgery: A retrospective review

The Efficacy of Routine Central Venous Monitoring in Major Head and Neck Surgery: A Retrospective Review Niels F. Jensen, MD, * Michael M. Todd, Robert I. Block, PhD,** Raymond L. Hegtvedt, CRNA,$ Timothy M. McCulloch, MD5 Department ty, IA 52246.

*Assistant

Professor

**Associate tProfessor $Staff thetist

of Anesthesia

of Anesthesia Certified

Registered

Nurse-Anes-

OAssistant Professor of OtolaryngologyHead and Neck Surgery Reprints

will

not be available.

Received for publication January 27, 1994; revised manuscript accepted for publication Anril 8, 1994.

Journal of Clinical Anesthesia 7: 119-125, 0 1995 by Elsevier Science Inc. 655 Avenue of the Americas, New York,

University

of

Iowa

Hospitals

and

Clinics,

Iowa

Ci-

Study Objective: To further define the efficacy of routine central venous catheter placement for major head and neck surgery from the standpoint of fluid and blood administration, and various other parameters of perioperative management. Design: Randomized, retrospective chart review. Setting: University-affiliated medical center. Patients: 104 patients who had undergone major head and neck surgery (defined as surgery lasting longer than 4 hours with a predicted blood loss of 500 ml or greater) at the University of Iowa Hospitals and Clinics between 1985 and 1992. Measurements and Main Results: Central venous monitoring was used in 51 of the 104 (49%) procedures. Patients with and without central monitors did not differ in age, weight, preoperative laboratory values [i.e., hemoglobin (Hb), blood urea nitrogen (BUN), creatinine), incidence of significant cardiac or renal disease, or a smoking history exceeding 30 pack years. In addition, these patients did not differ with respect to the following intraoperative characteristics: general type of anesthetic; duration of surgery; estimate of blood loss; Hb values; lowest urine output per hour; development of oliguria; total urine output; amount of replacement of blood, colloid, or crystalloid; development of systolic blood pressure less than 70 mmHg; or use of a myocutaneous flap. Patients also did not differ with respect to the following postoperative characteristics: duration of stay in the surgical intensive care unit or hospital, BUN or creatinine values on days 1 and 2, total urine output or the development of oliguria on days 1 through 3, incidence of reintubation, fever on days 1 through 5, wound dehiscence, death, myocardial infarction, or the development of pneumonia, pulmonary edema, or sepsis. Patients with central monitors had a greater incidence of having a tracheostomy performed and a slightly lower Hb level on the first postoperative o!ay than those without central monitors. Conclusions: The study raises doubt about the efficacy of routine central venous catheter placement as a necessary guide for fluid and blood administration for these procedures, or as a necessary adjunct for several other parameters of perioperative management. It suggests the need for a randomized, prospective evaluation.

of Anesthesia

Professor

of Anesthesia,

MD,?

Keywords:

Catheterization: central venous, monitoring of; surgery: head

and neck. 1995 NY

10010

SSDI

0952-8180/95/$10,00 0952~8180(94)00025-4

Original

Contribution

Introduction

Table

The efficacy of many medical and surgical interventions is unknown or is poorly understood.’ One example concerns central venous pressure (CVP) monitoring during major surgery. Central venous catheters are typically placed to assist in the assessment of cardiac function and intravascular volume status (as measured by right ventricular filling pressures), to provide secure intravenous (IV) access, and to allow vasoactive drugs to be delivered directly into the central circulation. A CVP catheter is usually used in major surgery in which large fluid shifts are expected, either acutely or over several hours. However, despite a wealth of information on the technical aspects of CVP catheter placement and the complications associated with it, there is little understanding of the degree to which the benefits of central line placement in routine surgery outweigh the risks (and costs). One example of a major procedure in which CVP monitoring is frequently employed is major head and neck surgery, particularly in patients with malignancies. Many surgeons and anesthesiologists feel that a CVP catheter will result in more rational, appropriate, and judicious fluid administration during these often lengthy cases. Better fluid and volume management should be associated with greater intraoperative and postoperative hemodynamic stability and fewer complications such as hypotension and oliguria, pulmonary edema, congestive heart failure (CHF), myocardial infarction (MI), and renal failure. The current report is a retrospective analysis of 104 surgical patients. Central venous catheters were placed at the descretion of the attending anesthesiologist. The data were reviewed to determine whether any evident management differences could be seen between those patients with and without CVP monitoring. This study represents our first approach to developing a clearer understanding of the role of routine central venous monitoring. It also serves as the pilot for a prospective study that is currently beginning. We hope that work such as this will permit a more rational use of interventions such as CVP monitoring.

Materials

and Methods

Between 1985 and 1992, approximately 63 1 patients underwent elective major head and neck surgery at either the University of Iowa Hospitals and Clinics or the Iowa City Veterans Administration Hospital. These adjacent hospitals are staffed by members of the sameanesthesia and otolaryngology departments. Major head and neck surgery was defined as a procedure with a minimum duration of four hours and an expected blood lossequal to or greater than 500 ml. The large majority of these casesinvolved head and neck cancer resection and reconstruction. A typical procedure was radical neck dissection(s) with a pectoralis major myocutaneous flap for resection of a pyriform sinus malignacy. The charts of 104 patients were randomly selected and were retrospectively reviewed by two of us (RLH, NFJ). Over 97% of the casesselected involved head and neck malignan120

J. Clin. Anesth., vol. 7, March 1995

Cancer

Anatomic Location of Lesions

1.

Location

Oral cavity

Pharynx (naso-,oro-, hypo-) Larynx (supraglottic,glottic, subglottic) Nose-sinus

Jaw

Salivary glands Neck

Thyroid-parathyroid Ear Other (non-cancer)

Percent of Total (n = 104)

5 47 24 2 6 6 6
ties. Charts were selected on the basisof the procedure performed; the use of a central venous catheter was not a criterion for review. No attempt was made prior to chart review to determine whether a patient had a central line placed as part of anesthetic monitoring. The anatomic location of lesionsin the study group is shown in Table 1. Specific data excerpted from each chart are listed below. Preoperative Variables Specific preoperative data recorded on each patient included age, weight, hemoglobin (Hb), blood urea nitrogen (BUN), and creatinine within one week prior to surgery. In addition, a significant history of cardiovascular disease(defined as chronic hypertension or a history of angina pectoris, a previous MI, or CHF) was noted, as was a history of significant renal disease (defined as a history of renal failure or a preoperative elevation of either BUN or creatinine). A history of cigarette smoking (more than 30 pack years) was also recorded. Intraoperative Variables Data extracted from the intraoperative records included primary anesthetic drugs used during the procedure; surgical duration from incision to final suture placement; estimated blood loss; lowest hourly intraoperative urine output in milliliters; presence of intraoperative oliguria (defined as a urine output of less than 0.5 ml/ kg/hr at one or more hourly intervals throughout surgery); total intraoperative urine output in milliliters; total intraoperative administration of blood, colloid, and crystalloid; the presence/incidence of one or more recorded systolic blood pressures (SBPs) less than 70 mmHg at any time following incision; need for a tracheostomy; and need for a myocutaneous flap to properly close the wound. Postoperative Variables Specific data from the postoperative period included total postsurgical intensive care unit (ICU) and hospital

Central venozcs monitoring

days; Hb on postoperative day one; BUN and creatinine on postoperative days one and two; total postoperative urine output from days 1 through 3; oliguria (any hourly urine output on postoperative days 1 through 3 of lessthan 0.5 ml/kg/hr); unplanned reintubation postoperatively; fever (defined as any postoperative recorded ICU or hospital temperature elevation greater than 38.5”C); wound dehiscence (abnormal early surgical wound breakdown); or death (during the postoperative period of hospitalization). In addition, we evaluated records for the occurrence of MI, pneumonia, pulmonary edema, and sepsis.We did not attempt to independently define criteria for these occurrences but rather relied on notations provided by the surgical and ICU teams and consultants. We made no attempt to judge the validity of the diagnosis made. For example, we did not establish rigid ECG or isoenzyme criteria for MI. Instead, the chart was reviewed to determine whether such a diagnosis had been made either by the surgical service and/or a consultant cardiologist.

Statistical Analyses Patients with and without central monitors were compared by t-tests for continuous variables such as age, and Fisher’s exact tests for dichotomous characteristics like tracheostomy. (Fisher’s exact tests were used becausethe low incidences of some characteristics, e.g., reintubation, made use of chi-square tests inappropriate.)

Results Anesthesia care for the patients was provided by 36 different anesthesiologists, 55 different residents, and 93 different combinations of anesthesiologists and residents. The 104 patients consisted of 96 different individuals, six of whom had two surgeries qualifying for inclusion and one of whom had three surgeries qualifying for inclusion. Multiple surgeries for these seven individuals were included in the analysis becausewe considered the combination of patient, anesthesiologist, and resident rather than the patient alone to be the appropriate “experimental unit” for evaluation. Central venous monitoring was used in 5 1 of the 104 (49%) procedures. Again, the use (or absence)of a central venous catheter was not a criterion for chart review and was not known to physicians or personnel in chart control who obtained the records prior to chart evaluation. Table 2 compares patients with and without central monitors with respect to preoperative variables [i.e., historical and other characteristics that could not have been affected by placement of a central monitor although they may have had some influence on the decisions to place a central catheter]. As indicated in Table2, patients with and without central monitors did not differ in age, weight, or preoperative laboratory values (Hb, BUN, creatinine). They also did not differ in incidence of sig-

in head and neck surgery:

Jensen et

al.

Table 2. Dependent,or PossiblyDependent, PreoperativeVariables Central Monitor (n = 51)*

Age (yr4 Weight(kg) Hemoglobin(g/dl) BUN (mgidl) Creatinine(mg/dl) Significantcardiac history (HTN, angina,CHF, MI) (%) Significantrenal history(%) Smoking(more than 30 pack-yrs) (%)

62.6 + 14.6

69.0 f 15.9 13.6 k 1.7 14.6 k 7.4 1.1 t 0.2

No Central Monitor (n = 53)*

57.6 k 73.6 2 14.0 * 13.9 k 1.1 f

51

53

14

10

51

57

12.8 18.5 2.0 5.6 0.3

Note: Dataother than percentages are means-+ standarddeviation. BUN = bloodureanitrogen;HTN = hypertension;CHF = congestiveheartfailure; MI = myocardialinfarction. *The percentages of missingdata werelessthan 2% for all variables.

nificant cardiac or renal disease, or a smoking history exceeding 30 pack years. Table 3 compares patients with and without central monitors with respect to dependent, or possibly dependent, variables intraoperatively (i.e., characteristics that might have been affected by placement of a central monitor during the intraoperative course). Patients with and without central monitors did not differ with respect to the following intraoperative characteristics: general type of anesthetic; duration of surgery; estimate of blood loss; hemoglobin values; lowest urine output per hour; development of oliguria; total urine output; amount of replacement of blood, colloid, or crystalloid; development of a SBP lessthan 70 mmHg; or use of a myocutaneous flap. Patients with central monitors had a higher incidence of tracheostomy compared with those without central monitors. Table 4 compares patients with and without central monitors with respect to dependent, or possibly dependent, variables postoperatively. Patients did not differ with respect to the following postoperative characteristics: duration of stay in the surgical ICU or hospital, BUN or creatinine values on days ! and 2, total urine output or the development of oliguria on days 1 through 3, incidence of reintubation, fever on days 1 through 5, wound dehiscence, death, MI, or the development of pneumonia, pulmonary edema, or sepsis.Patients with central monitors had slightly lower Hb levels on the first postoperative day than those without central monitors. Since length of stay in the surgical ICU and in the hospital after surgery might conceivably have varied between the University Hospital and the Veteran’s Administration Medical Center, and such variations might have J. Clin. Anesth.,vol. 7, March 1995

121

Original

Contributions

3. Dependent, or Possibly Dependent, Intraoperative Variables Table

Central Monitor (n = 51)*

Nitrous oxide-isoflurane-fentanyl anesthetic (%) 84 Duration (min) 659.5 ” 188.6 Blood loss (ml) 1,375.g f 1,785.2 Hemoglobin (g/dl) 9.9 2 1.9 Lowest urine output per hr 32.5 + 17.3 04 Oliguria (less than 0.5 ml/kg/hr) (%) 51 Total urine output 1,189.7 ? 735.5 (ml) Replacement blood 415.7 f 683.3 04 Replacement colloid (ml) 966.2 ? 708.1 Replacement crystalloid (ml) 5,723.g f 1,730.5 Systolic blood pressure less than 70 mmHg 8 (%) Tracheostomy (%) 93 Myocutaneous flap 22 (%)

Table

4.

Dependent,or PossiblyDependent,

Postoperative

No Central

Variables

Monitor

Central Monitor (n = 51)*

(n = 53)*

91 635.7 + 244.3 893.6 ” 667.1 10.2 2 1.9 29.8 t 15.3 57 1,089.g If: 824.9 253.8 k 378.1 1,038.5 k 776.8 5,218.l f 2,158.0

0

75 21

Note: Data other than percentages are mean f standard deviations. *The percentages of missing data were less than 2% for all variables. tDiffers from patients without a central monitor, by Fisher’s exact test, p < 0.05.

obscured effects of central monitors on these variables, separate analysesof these variables were conducted for each institution. At neither institution did patients with and without central monitors differ in lengths of staysin the surgical ICU or in the hospital after surgery.

Surgical intensive care unit stay (days) Hospital stay (days) Hemoglobin, day 1 WV BUN, day 1 (gidl) BUN, day 2 Creatinine, day 1 (g/4 Creatinine, day 2 Total urine output (UOP), days l-3 (ml) Oliguria (UOP ~0.5 mg/kg/hr) (%) Reintubation (%) Fever postoperative days (POD) l-5 (%I Wound dehiscence (%I Death (%) Myocardial infarction POD l-5 (%) Pneumonia (%) Pulmonary edema (%) Sepsis (%)

No Central Monitor (n = 53)*

3.3 f 5.3 26.7 2 28.3

3.4 2 7.1 22.3 * 21.3

10.0 2 1.4t 12.0 + 5.4 9.0 * 4.1

10.8 k 1.6 11.7 k 4.9 9.1 t 3.8

0.9 2 0.2 0.8 2 0.2

1.0 * 0.2 0.9 f 0.2

5,560.l 2 2,075.l

5,743.5 f 1,919.5

26 0

20 6

35

31

14 4

8 4

0 12

0 13

10 10

6 4

Note: Data other than percentages are means * standard deviations. *The data for replacement colloid are based on the 58% of patients to whom it was administered. The percentages of missing data are 19% for intraoperative hemoglobin, 17% for blood urea nitrogen (BUN) and creatinine on postoperative day 2, 21% for total urine output on postoperative days 1 to 3, 12% for postoperative oliguria, and less than 5% for the other variables. tDiffers from patients without a central monitor, t(lOO) = 2.8, p < 0.01.

Discussion CVP is the blood pressure at the junction of the vena cavae and the right atrium. As such, it is a measure of right ventricular filling pressure and, hence, an indirect measure of the function of the right atrium and right ventricle. According to Otto,* the most common reason for CVP measurement is to “determine the state of cardiac function and the state of vascular volume.” There are many other indications for central venous access,the most important for the anesthesiologist include the rapid administration of fluids and drugs, the definitive venous access,and its aid in operations where the risk of venous air embolism is particularly high (such as the sitting position).

122

J. Clin. Anesth., vol. 7, March 1995

While there has been great attention paid to timing, 3,4 insertion sites and techniques,5S616and complications1746of central venous monitors, there is little information regarding the efficacy of the monitor. In his 1983 Rovenstine Lecture to the American Society of Anesthesiologists,Dr. Arthur Keats objected to the use of technology implemented for logical but not necessarily scientific criteria. 48A decade later, there is still no shortage of “how to” studies but there remains a critical shortage of “whether to” studies.4s Our goal in undertaking this study was to determine whether the presence of central venous monitoring in major head and neck surgery significantly effects peri-

Central

operative management, especially as it relates to the administration of fluids, blood, and intraoperative hemodynamics. Before discussing our results, the major limitation of this study must be addressed: its retrospective nature. For the study to be valid, it is crucial that the two groups be similar except for the variable tested: the presence or absence of a central venous catheter. If the groups were not similar and central venous monitoring was performed on a selected group, then subsequent management might be due to factors totally unrelated to central venous monitoring and invalidate these results. Therefore, the issue of bias in patient selection is especially crucial. What is the evidence for bias in group selection, for unintended differences in group selection and/or composition? The only preoperative, intraoperative, or postoperative variables that distinguished patient groups with and without central monitors were the incidences of tracheostomy, which was higher in patients with central monitors than those without central monitors, and Hb levels on the first postoperative day, which were lower in patients with central monitors than those without central monitors. While this may be a Type 1 statistical error, it is conceivable that a planned tracheostomy somehow affected the anesthesiologist’s perception of the magnitude of the forthcoming surgery and, therefore, increased the probability of a decision to place a central line. However, in reviewing the surgical procedures (planned and actually performed) in the tracheostomy group, we found it difficult to believe that they were substantially different. For example, when the 14 patients who did not undergo tracheostomy were deleted from the analysis and the impact of central catheter placement again examined, we found no differences in any variable. Estimated blood loss is another area suggesting possible bias. Estimated blood loss in the CVP group, while not sigGfcantl~ greater, was somewhat greater. It is possible that, by altering the anesthesiologist’s perception of the severity of hemodynamic changes or the magnitude of the operation, the placement of a CVP catheter might be associated with an overestimation of blood loss. In other words, the anesthesiologist in these cases expected a larger blood loss, placed a central venous catheter to monitor and correct it, and then erroneously overestimated blood loss. More careful attention to and standardization of the intervals between recording blood loss as well as the means of blood loss determination in a prospective evaluation would clarify this issue. For example, carefully weighing sponges at defined intervals rather than just casually glancing at them periodically may improve the accuracy of blood loss estimation. Nevertheless, the differences in blood loss between the two groups were small, and, in our opinion, is unlikely to be related to CVP monitoring. As the data show, very few major management differences (or management consequences) existed between the two groups in our study. In addition, the large number of different personnel involved in the care of these patients, combined with the

venous monitoring

in head and neck surge?:

Jensen et al.

lack of any widespread guidelines for CVP monitoring suggests that our data are not confounded by any major systematic bias. It appears that the two groups are comparable. However, the issue of bias can only be definitively resolved with a prospective study, which is now starting. With the above comments in mind, this retrospective analysis suggests the following: (1) the decision to employ CVP monitoring for major head and neck surgery is not associated with major differences in intraoperative fluid management. (2) There was no difference in the two groups in the frequency of several intraoperative and postoperative complications that may be related to fluid administration, including pulmonary edema, oliguria, and wound dehiscence. (3) There was no difference in the two groups in total number of surgical ICU and hospital days. Invasive medical practices, which are often expensive and not without risk, should rest on a firm foundation of proven efficacy. 5o Unfortuna t e1y, the determination and scientific foundations of efficacy in many, if not most, health care areas is simply not known. According to the Institute of Medicine, valid randomized; controlled efficacy studies have been applied to only a very small part of medical practice. lS5’ Wennberg5’ partly explains the absence of firm evidence of efficacy as the two- to threefold variation in the use of some medical and surgical procedures between various regions. In these cases, he suggests, expert medical opinion has no choice but to accept as appropriate a broad set of indications for treatment.52 Eddy’ examined many well accepted conventional treatments to learn if valid efficacy studies exist regarding recommended practice or alternative managements. He concluded that virtually none have been evaluated with well-designed, controlled studies comparing alternative interventions. It appears that many “standard” diagnostic and therapeutic practices, involving huge numbers of patients, high risk, and tremendous cost, rest upon very uncertain foundations with respect to efficacy. lx’* Our study suggests that this may apply to CVP monitoring in major head and neck surgery. Despite the shortcomings noted above, to our knowledge, this retrospective review represents one of the first attempts to evaluate the efficacy of a very common invasive procedure, the routine placement of central monitoring for major head and neck surgery. The drawbacks and limitations are important but the implications of our retrospective study are troubling. It suggests that, in this surgical population, the placement of a central venous monitor may contribute little to anesthetic management. The randomness of placement suggests that anesthesiologists may lack firm criteria for the use of this monitor. Nevertheless, this conclusion must be viewed with caution. The sample population is not large, and it represents only one general type of surgical procedure. It is possible that in some other patient group, which underwent a different type of surgical procedure, a different result might have been seen. However, we suggest that

J. Clin.

Anesth.,

vol.

7, March

1995

123

Original

Contributions

the data do indicate the need for valid prospective evaluation of the indications (and efficacy) of CVP monitoring in surgery.

20.

References

21.

1. Eddy D: Uncertainty, outcomes, and the quality of medical care. In: Joint Commission on Accreditation of Healthcare Organizations (JCAHO): Cornerstones of Health Care in the Nineties: Forging a Framework of Excellence. Oakbrook Terrace, IL: JCAHO, 1990: 17. monitoring. In: Blitt C (ed): 2. Otto CW: Central venous pressure Monitoring in Anesthesia and Critical Care Medicine, New York: Churchill Livingstone, 1990: 169-210. 3. Dzelzkalns R, Stanley TH: Placement of the pulmonary arterial catheter before anesthesia for cardiac surgery: a stressful, painful, unnecessary crutch. J Clin Monit 1985; 1: 197-200. 4. Streisand JB, Clark NJ, Pace NL: Pulmonary arterial catheterization before anesthesia in patients undergoing cardiac surgery. Placement of the pulmonary arterial catheter before anesthesia for cardiac surgery: safe, intelligent, and appropriate use of invasive hemodynamic monitoring. J Clin Monit 1985; 1:193-7. 5 Sanford TJ Jr: Internal jugular vein cannulation versus vein cannulation. An anesthesiologist’s view: the right internal jugular vein. J Clin Monit 1985; 1:58-60. 6. Horrow JC, Metz S, Thickman D, Frederic MW: Prior carotid surgery does not affect the reliability of landmarks for location of the internal jugular vein. Anesth Analg 1987;66:452-6. 7. Hoyt DB: Internal jugular vein cannulation versus subclavian vein cannulation. A surgeon’s view: the subclavian vein.J Clin Monit 1985;1:61-3. 8. Oda M, Fukushima Y, Hirota T, Tanaka A, Aono M, Sato T: Forum. The para-carotid approach for internal jugular catheterization. Anaesthesia 1981;36:896-900. 9. Tyden H: Cannulation of the internal jugular vein-500 cases. Acta Anaesthesiol &and 1982;26:485-8. 10. Webre DR, Arens JF: Use of cephalic and basilic veins for introduction of central venous catheters. Anesthesiology 1973; 38:389-92. CL, Wright WA, Otto CW: J-wire versus 11. Blitt CD, Carlson straight wire for central venous system cannulation via the external jugular vein. Anesth Analg 1982;61:536-7. JR: A safe way to perform infraclavicular 12. Borja AR, Hinhaw subclavian vein catheterization. Surg Gynecol Obstet 1970; 130: 673-6. 13. English IC, Frew RM, Pigott JF, Zaki M: Percutaneous catheterization of the internal jugular vein. Anaesthesia 1969;24: 521-31. of central vascular catheters [Letter]. N 14. Fabri PJ: Replacement Engl J Med 1993;328:445. 15. Nordstrom L, Fletcher R: Comparison of two different J-wires for central venous cannulation via the external jugular vein. Anesth Analg 1983;62:365. 16. Pina J, Morujao N, Castro-Tavares J: Internal jugular catheterisation. Blood reflux is not a reliable sign in patients with thoracic trauma. Anaesthesia 1992;47:30-1. TC, Tan WT: A central venous catheter complicating 17. LIM head and neck surgery [Letter]. BrJ Oral Maxillofac Surg 1992; 30:202. RL, Johnson WE, Gravlee GP, Brauer S, Richards D: 18. Royster Arrhythmias during venous cannulation prior to pulmonary artery catheter insertion. Anesth An&g 1985;64: 1214-6. 19. Aoki H, Mizobe T, Nozuchi S, Hatanaka T, Tanaka Y: A lethal

124

J. Clin.

Anesth.,

vol. 7, March

1995

complication

22.

23. 24. 25. 26. 27. 28. 29. 30.

31.

32. 33.

34. 35. 36.

37. 38.

39. 40.

4 1. 42. 43. 44. 45.

46.

of internal jugular vein catheterization [Abstract]. Anesth Analg 1993;76:S910. Blackshear RH, Gravenstein N: More on risk factors for central venous catheter-related vascular erosions [Letter]. J Parenter Enteral Nutr 1992;16:595-6. Breznick DA, Ness WC: Acute arterial insufficiency of the upper extremity after central venous cannulation. Anesthesiology 1993;78:594-6. Camilleri AE, Davies FW. Aberrant central venous catheter complicating radical neck dissection. J Laryngol Otol 1991; 105: 491-2. Carey BE: Major complications of central lines in neonates. Neonatal Netw 1989;7: 17-28. Chou NK, Wu YL: Unusual complication of central venous catheterization [Letter]. AmJ Emerg Med 1993;11:189-91. Glance KM, Glauser FL: Sudden apnea following attempted central line placement. Chest 1991;100:505-6. Collins JL: Central venous catheter complications [Letter]. Onco1 Nurs Forum 1991;18:819-20. Dawood MM, Trebbin WM: Complications associated with central venous cannulation. Hasp Prac 1991;26:21 l-4, 218-9. Doblar DD, Boyd GL: Air embolism during attempted central line placement [Letter]. Chest 1992;102:1917. Fitchie C: Central venous catheter-related infection and dressing type. Intensive Crit Care Nurs 1992;8: 199-202. Goutail-Flaud MF, Sfez M, Berg A, et al: Central venous catheter-related complications in newborns and infants: a 587-case survey. J Pediatr Surg 1991;26:645-50. Hruszkewycz V, Holtrop PC, Batton DC, Morden RS, Gibson P, Band JD: Complications associated with central venous catheters inserted in critically ill neonates. Infect Control Hasp Epidemiol 1991;12:544-8. Johnson A, Oppenheim BA: Vascular catheter-related sepsis: diagnosis and prevention. J Hasp Infect 1992;20:67-78. Joshi P, Bullingham A, Soni N: Septic atria1 thrombus. A complication of central venous catheterisation. Anaesthesia 1991;46: 1030-2. Kimmel RD, Swiacki CR: Catheter complications [Letter]. Chest 1990;97:1271-2. Krasna IH, Krause T: Life-threatening fluid extravasation of central venous catheters. J Pediatr Surg 1991;26: 13468. McGurk M, Thomson PJ: A central venous catheter complieating head and neck surgery. Br J Oral Maxillofac Surg 1991; 29:388-91. Michel MZ, Ryder I: Potential hazard of central venous cannulation [Letter]. Anaesthesia 1991;46:156. Mukau L, Talamini MA, Sitzmann JV: Risk factors for central venous catheter-related vascular erosions. J Parenter Enteral Nutr 1991;15:513-6. Parsa MH: Complications of central venous catheter insertion [Letter]. Crit Care Med 1992;20:443-4. Polos PC, Sahn SA: Complication of central venous catheter insertion: fragmentation of a guidewire with pulmonary artery embolism. Crit Care Med 1991;19:438-40. Powell H, Beechey AP: Internal jugular catheterisation. Case report of a potentially fatal hazard. Anaesthesia 1990;45:458-9. De Silva P, Earl U, Sinha S: Complications of a central venous line in the newborn [Letter]. Arch Dis Child 1990;65:915. Smith MB: Another complication of intravenous catheterization: a case report. J Okla State Med Assoc 1990;83: 18-20. Toltzis P, Goldmann DA: Current issues in central venous catheter infection. Ann Rev Med 1990;41:169-76. Woodyard TC, Mellinger JD, Vann KG, Nisenbaum J: Acute superior vena cava syndrome after central venous catheter placement. Cancer 1993;71:2621-3. Yerdel MA, Karayalcin K, Bozatli L, Ertunc T: An unusual

Central

outcome of a common complication of the subclavian vein catheterization [Letter]. Crit Care Med 1990;18:7967. 47. Keats AS: The Rovenstine lecture, 1983: cardiovascular anesthesia: perceptions and perspectives. Anesthesiology 1984;60: 467-74. 48. Tinker JH, Dull DL, CaplanRA, Ward RJ, CheneyFW: Role of monitoringdevicesin preventionof anestheticmishaps:a closedclaimsanalysis. Anesthesiology 1989;7 1:541-6. 49. JensenN, Tinker JH: Quality in medicalcare: lessons from

Effects of the Benzodiazepine Performance Following Total

Department

Venous monitoring

in head and neck surgery: Jensen et al.

industry and a proposalfor valid measurement and improvement.Qua1Perform Qua1 Health Care1993;1:138-53. 50. Institute of Medicine:Assessing Medical Technologies. Washington, DC: NationalAcademyPress,1985. 51. WennbergJE: The paradoxof appropriatecare.JAMA 1987; 258:2568-g. 52

Antagonist Intravenous and Alfentanil

Bunker JP: Is efficacy the gold standardfor quality assessment?Inquiry 1988;25:51-8.

Flumazenil on Postoperative Anaesthesia with Midazolam

A. Nilson, M.P. Person, and P. Hartvig of Anaesthesiology, Uppsala University Uppsala, Sweden

Hospital,

Abstract Postoperative performance following total intravenous anesthesia(TIVA) using midazolam and alfentanil was studied with and without the administration of a single dose of a benzodiazepine antagonist, flumazenil (Ro 15-1788). Performance was compared with a reference group anaesthetized with thiopentone, alfentanil and nitrous oxide. All patients were assessedby use of a rating scale which took into account the degree of sedation, amnesia, comprehension and cooperation as well as temporal and spatial orientation. There was a slow recovery following TIVA with somnolence and amnesia lasting several hours. Administration of flumazenil 1.O mg i.v. at extubation caused a significant reduction of sedation (P < 0.001) during the first postoperative hour, with patients fully awake or only lightly sedated, but was later followed by resedation. The patients of the reference group were moderately sedated during the observational period. Five and six hours postoperatively there was no difference between the groups. Amnesia was more profound in the groups that received midazolam; the effect of the antagonist could only be seen for 15 min after its administration. Comprehension and cooperation, as well as orientation, were equally good in the antagonist and in the reference group during the immediate postoperative period, whereas in the TWA group a gradual improvement over the first hours was seen. In the antagonist group there wasno increase in the number of analgesic requirements, no anxiety attacks or other adverse effects. It is concluded that flumazenil offers an improvement in postoperative performance following TIVA induced by midazolam and alfentanil, but the effects are of short duration. Reprinted with permission from Acta Anaesthesiologica

Scandinavica

1988;32:441-6.

J. Clin. Anesth.,vol. 7, March 1995

125