Should surgeons still be implanting pacemakers?

Should Surgeons Still Be Implanting Pacemakers? T. Bruce Ferguson, Jr, MD, Bruce D. Lindsay, MD, and John P. Boineau, MD Division of Cardiothoracic Surgery, Department of Surgery, and Section of Electrophysiology, Division of Cardiology, Department of Medicine, Washington University School of Medicine, St. Louis, Missouri

Cardiac pacing has undergone major changes in the areas of manpower, technology, and cost over the past 10 years. Arguments have been made to eliminate cardiac surgical involvement in pacing on the basis of these three areas of change: implantations are increasingly performed by nonsurgeons, surgeons have not kept up with the technologic advances in pacing, and consolidation of bradypacing resources is necessary during a time when reimbursement has declined significantly. This study examined two eras of pacing therapy at an institution where pacemaker implantation has always been performed by cardiothoracic surgeons. The purpose of the study was to critically analyze (1) the current role (if any) of cardiothoracic surgeons in delivery of pacemaker therapy and (2) the current results of cardiothoracic surgical involvement in pacemaker implantation. In 1,562 procedures performed between 1986 and 1992, the infection rate was 0.51% and the overall complication rate (both short-term and long-term) was 5.2%. During

T

he field of cardiac pacing has changed dramatically in three areas since the placement, in 1962 by a surgeon, of the first wholly implantable pacemaker system. The first area of change has been manpower; for the majority of the 32-year interval since 1962, the implantation of pacing devices has been performed by individuals (1) with specific training in the techniques of surgery, either general surgery or cardiovascular surgery, and (2) with the capability of handling the complications of the procedure(s). As with many other areas of cardiovascular surgical disease, however, the impetus to keep pacemaker implantation under the umbrella of surgery diminished as more financially rewarding procedures increased in volume in many practices. Internists, invasive cardiologists, and cardiologists with specific expertise in electrophysiology have gradually filled the void, such that by the early part of this decade approximately 50% of bradycardia therapy devices are implanted by nonsurgeons [ 11. The second area of change has been in pacing technology. The complexities of dual-chamber rate-responsive pacing are much more difficult to master than the simple, nonprogrammable single-chamber devices used for many years; furthermore, the rapidity with which new technoPresented at the Fortieth Annual Meeting of the Southern Thoracic Surgical Association, Panama City Beach, FL, Nov 4-6, 1993. Address reprint requests to Dr Ferguson, Suite 3108 Queeny Tower, Barnes Hospital Plaza, St. Louis MO 63110.

0 1994 by The Society of Thoracic Surgeons

era 1 (1/1/86 to 6/30/89), 80% of implants were singlechamber and follow-up was incomplete and dependent in many instances on the referring cardiologist/internist. For the implantations performed in the second era (7/1/89 to 12/31/92) as part of an established Pacemaker Service, complete clinical and transtelephonic follow-up services were provided by this coordinated medical-surgical approach. During era 2, 53.9% of implants were dualchamber (79% during 1992). Total and infectious complication rates remained low in era 2 despite this change in technology. These data mitigate against the arguments made above to eliminate surgical involvement in bradypacing, and demonstrate that cardiothoracic surgical expertise can be integrated into and significantly contribute to a ”Center of Excellence” approach to pacemaker therapy. These issues should be considered as government, third-party payors, and hospitals consider bradypacing therapy as a target for health care reform. (Ann Thorac Surg 1994;57:588-97)

logic advances are being made makes it difficult to keep abreast of current techniques in pacing in the absence of a demonstrated specific interest in this field. Finally, increases in the costs associated with pacing have occurred as more high-cost technology is being used and newer indications for pacemaker therapy are defined. Associated with this, however, has been the implementation of Resource-Based Relative Value Scale guidelines for pacemaker implantation, often at 1996 reimbursement rates, which has lessened the financial incentive to be involved in cardiac pacing, regardless of one’s specialty. In addition, the overall pressures to limit in-hospital costs, and the difficulties in separating implantation costs from other inpatient medical costs, further complicates these issues in this predominantly Medicare-funded patient population. These changes in manpower, technology, and cost have made it difficult for individuals and organizations outside of medicine to assess the effectiveness of bradypacing therapy, which is now perceived as an attractive target for healthcare reform. The assessment of cardiothoracic surgical involvement in pacing is that of low interest and inadequate expertise; one way to “simplify” pacing therapy would be to eliminate surgeons from the therapeutic

loop. It appears valid, then, to question whether surgeons, and specifically cardiothoracic surgeons, should continue to be involved in device therapy for bradyarrhythmias. If the answer to this question is yes, then the nature and 0003-4975/94/$7.00

FERGUSON ET AL SURGEONS AND CARDIAC PACING

Ann Thorac Surg 1994;57:588-97

extent of this involvement should perhaps be more clearly elucidated in light of these changes. The purpose of this study was to examine the practice of cardiac bradycardia pacing at an institution where a large volume of pacemaker implantations have been performed exclusively by cardiothoracic surgeons since the inception of pacemaker therapy. Two "eras" are analyzed: the first, from 1/86 until 7/89, encompasses an interval when pacemaker implantations were performed by cardiothoracic surgeons with many years of experience in pacemaker implantation but no formal training or expertise in electrophysiology, and during which time no formal follow-up mechanisms were available; and the second, from 7/89 through 12/92, encompasses an interval when implantations were performed by a single cardiothoracic surgeon with a specific interest in surgical electrophysiologic problems and when the implantations were performed as part of an established Pacemaker Service.

Material and Methods From 1/86 until 12/92, 1,562 total pacemaker procedures were performed at Barnes Hospital at the Washington University School of Medicine in St. Louis. All procedures were performed by board-certified cardiothoracic surgeons in the operating room suite under surgical aseptic technique. Portable fluoroscopic x-ray support was provided by the Mallinckrodt Institute of Radiology at Washington University. Except when non-conduction-system cardiovascular disease dictated otherwise, all patients were operated on under local 1% lidocaine anesthesia, with anesthesia coverage standby as indicated by preexisting medical conditions.

Era 1 Between 1/86 and 6/89,576 procedures were performed by two surgeons who had been involved with cardiac pacing since its inception. Routine intravenous (three doses) and oral antibiotic therapy (1 week postoperatively) was administered to all patients. Surgical cutdown techniques on the cephalic vein, the external jugular vein, or both, were used in more than 95% of the implantations. Passivefixation ventricular leads were implanted in all patients; atrial leads, when placed, were evenly divided between active-fixation and passive-fixation types. During this interval, a predominant number of generator devices were supplied by a single vendor. The clinical decision regarding the type of pacing system (single versus dual chamber) to be implanted was most often made by the referring cardiologist. Technical support for the implants was most commonly provided to the surgeon by an individual working for Barnes Hospital with acquired expertise in pacemaker technology; industry technical support personnel were used for the implants when the Barnes individual was unavailable. For most of this era, no organized pacemaker follow-up procedures were in place at Barnes Hospital or Washington University. Clinical follow-up was provided for the patients by the referring cardiologist in the large majority of instances. Beginning in 1986, a telephone follow-up

589

database service that had previously been established at another institution was brought to the Division of Cardiothoracic Surgery and began enrolling a limited number of patients. Surgical involvement in follow-up was limited to referrals for complications of implantation.

Era 2 In July 1989 the surgeon performing the majority of the implantations decided to retire, and the decision was made to bring bradypacing in a formal way back under the "umbrella" of surgical electrophysiology. The period from 7/89 through 12/92 constituted the second interval (era 2). The Washington University Pacemaker Service was established, and the deficiencies in follow-up that had existed in the past were addressed. A cardiologist with a joint appointment in Surgery and Medicine became the Director, and a cardiothoracic surgeon interested in electrophysiology and a cardiac electrophysiologist with experience in pacing were appointed as Co-Directors. Two pacemaker technicians, one primarily for the implant procedures and one primarily for clinical follow-up were hired and trained; cross-coverage between these two individuals eliminated the need for industry representatives except for devices implanted under Food and Drug Administration investigation protocols. In 1989, the Paceart (Paceart Inc, Wayne, NJ) pacemaker follow-up system was acquired for clinical and telephone follow-up monitoring; the existing telephone database was incorporated into this system and the telephone follow-up service was expanded to cover all new implants. By 1/91 essentially 100% of follow-up was being actively provided by the Pacemaker Service, through telephone follow-up and clinic follow-up according to existing Medicare guidelines. All implantations during this second era were performed in the operating room using portable fluoroscopic equipment, as mentioned earlier. Perioperative intravenous (three doses) antibiotics were administered, and the subclavian technique for lead introduction was used; cephalic or subclavian cutdown approaches or both were used only if the subclavian approach failed. The final decision regarding the type of system implanted was made by the surgeon, often in consultation with the referring physician and the other Co-Directors involved in the Pacemaker Service when indicated. Active-fixation ventricular and atrial leads were implanted in all patients. Because the subclavian technique was used for lead placement, radiographically documented pneumothorax was a potential complication during the second era that was not present during the first era. A portable chest roentgenogram was obtained on all patients immediately after implantation and the next morning after implantation; to minimize postimplantation hospital stay aggressive use of tube thoracostomy was employed. If no air leak was present on the morning after tube placement, the chest tube was removed. Postimplantation pacemaker interrogation and reprogramming was performed on the morning after implantation in the Pacemaker Clinic. A deliberate decision was made by the Directors to support three pacing hardware manufacturers and their

590

FERGUSON ET AL SURGEONS AND CARDIAC PACING

products to provide (1) access to a complete spectrum of available bradypacing products, ( 2 ) a breadth of training experience, and (3) protection for the Clinic from the negative impact of device failure and recall.

Data Analysis To assess implant risks and complications, the following definitions were used in this study: total implant procedures = all pacemaker-related procedures; total new implant procedures = all endocardial and epicardial single- and dual-chamber pacemaker systems implanted into patients without preexisting pacemaker systems; total lead reposition procedures = all reposition procedures, both for leads implanted at Barnes Hospital and for leads implanted in patients transferred from outside facilities for lead-related problems; lead reposition procedures (inhouse) = lead repositions after placement of leads at Barnes Hospital, regardless of interval between implantation and reposition; lead change procedcres = exchange of a lead or leads due to conductor fracture, insulation break, or recall, without evidence of system infection, with new implantation; total infected explant procedures = all patients undergoing explantation of a lead or generator due to infection, both in-house and outside implants; infected explant procedures (in-house) = explantation of systems implanted at Barnes Hospital that became infected, regardless of the interval from system placement to removal; hematoma evacuation procedures = hematoma evacuation in the perioperative period (no infection); noninfected procedures = generator pocket revision for impending erosion/discomfort/cosmesis (no evidence of local or systemic infection), or explantation of pacing systems after orthotopic cardiac transplantation; and total number endocardial leads implanted = number of single-chamber systems + 2 (number of dual-chamber systems) + number of lead changes. All demographic, implantation and perioperative data were obtained from the Division of Cardiothoracic Surgery Patient Database and the Paceart Database; additional data were obtained from the medical records when necessary. Statistical analyses were performed in consultation with the statistician in the Cardiothoracic Surgical Research Laboratory and with the Division of Biostatistics

c

300

B ez

p

8

0

z

MOthcr

250 200

0 Explants

€4 Ld Repos/A

0 Batt A

150

New Implant

100

50

0

1986

1987

1988

1989

1990

1991

1992

Fig 1 . Complete pacemaker experience, by year, at Barnes Hospital at Washington University, 1/1/1986 through 12/31/1992. (Batt A = battery change; Explants = explantation of infected pacemaker hardware; Ld ReposIA = lead reposition or lead change; Other = other surgical procedures, eg, noninfected explantation of preexisting pacemaker system after orthotopic cardiac transplantation.)

Table 1. T h e Barnes Hospital Experience (1986 to 1992)

Procedure Total procedures Total new

Number

1,562 1,268 (81.2%of total procedures)

implantations Transvenous SC 758 (59.8%of new implantations) Transvenous DC 474 (37.4%of new implantations) Epicardial SC 16 (1.3%of new implantations) Epicardial DC 20 (1.6%of new implantations) Battery changes 153 (9.8%of total procedures) Repositioned IH leads 53 (3.1%of endocardial leads placed) Periop repositionings 47 (2.7% of endocardial leads placed) Outside repositionings 18 (25.4%of leads repositioned) Lead changes 17 (1.0% of endocardial leads placed) Infected IH 8 (0.5% of total procedures) explantations Infected outside 3 (27.3%of all infected explantations explantations) 42 (2.7% of total procedures) Noninfected procedures DC

=

dual chamber;

IH

=

in-house;

SC

=

single chamber.

at Washington University School of Medicine, using Macintosh-based Systat (Statistical Analysis Systems, Deerfield, IL) software. Financial data are expressed as mean k standard deviation; era 1 and era 2 were compared using unpaired Student’s t test and ,$ analyses.

Results Total Experience, 1986 to 1992 Between 1/86 and 12/92, 1,562 total bradycardia pacemaker procedures were performed (Fig 1). Not included in this experience are procedures related to implantation of antitachycardia devices. Of the 1,562 procedures, 1,268 were new device implantations, 153 were generator changes for end-of-life indications, 11 were explantations of infected pacing systems, 3 were reexploration procedures for noninfected hematoma evacuation, 71 were for lead repositioning for loss of adequate pacing or sensing capability (both perioperative and late), 17 were lead changes for recall or insulation or conductor fracture, and 39 were explantation or revision procedures for late impending erosion or generator migration, or explantation of a preexisting pacing system after orthotopic cardiac transplantation (Table 1). There were no operative deaths; 2 patients required emergency thoracotomy for right ventricular perforation and cardiac tamponade, 1 in each era. The number of new implants per year is shown in Figure 2; the year 1989 is split, reflecting the establishment of the Pacemaker Service in July of that year. For the entire series, the in-house explantation rate for infectious complications was 0.51% (8/1,562 total procedures); two of these infectious complications were in a patient who was subsequently diagnosed with Munchausen’s disease. The in-house lead reposition rate was 3.1% (53/1,724 new leads implanted). The total complication rate (in-house lead reposition procedures, in-house infected explant procedures, hematoma evacuation pro-

FERGUSON ET AL

Ann Thorac Surg

SURGEONS AND CARDIAC PACING

1994;5758697

591

Table 2. Experience Between January 1986 and June 1989

(Era 1)

Procedure

1986

1987

1988

1989

1990

1991

992

Fig 2. New implant procedures during each year between 1986 and 1992. The year 1989 is split, reflecting the establishment of the Washington University Pacemaker Service in july of that year. See text for description of eras 1 and 2 .

cedures, pneumothorax, and exploration for tamponade) was 5.1% (80/1562).

Era 1 During era 1, the number of implants and relative percentages of implant type remained fairly constant (Fig 3). Of new transvenous implants, a total of 450 singlechamber and 92 dual-chamber implants were placed, and single-chamber implants constituted 79.3 to 80.5% of new implants during 1986 to 1988 (Table 2). The overall lead reposition rate for loss of pacing or sensing as a percentage of new implants was 3.9%; included in this group were 5 patients transferred from outside hospitals after lead dislodgements for further therapy at Barnes Hospital. The lead reposition rates for the years 1986 to 1988 after in-house implantations during era 1 were 2.2%, 2.6%, and 4.0% respectively. The lead reposition rates (as a percentage of the total number of new leads implanted in-house) during 1986 to 1988 were 2.0%, 2.3%, and 3.4% respectively; the total for era 1 was 2.7% expressed as a percent of new endocardial leads implanted. Of the 17 lead repositions during this 3.5-year interval, two were performed at an interval greater than 45 days after implantation. During era 1, a total of three explant/lead extraction procedures were performed for infectious complications. All three were infectious complications resulting from

900 800 700 600 500 ‘Z 400 300 200 100 yI

5

U

E 5 5 =

n

Dual-Chamber 0 Battery b

Other

ERA 1

ERA 2

Fig 3 . Distribution of procedures by pacemaker system type. (Other = lead reposition or change, explantations for infections, or other procedures on noninfected patients.)

Number

687 Total procedures 559 (81.4% of total procedures) Total new implantations 450 (80.5% of new implantations) Transvenous SC 92 (16.5% of new implantations) Transvenous DC 5 (0.9% of new implantations) Epicardial SC 12 (2.1% of new implantations) Epicardial DC 74 (10.8% of total procedures) Battery changes 17 (2.7% of endocardial leads placed) Repositioned IH leads Periop repositionings 15 (2.3% of endocardial leads placed) 5 (22.7% of leads repositioned) Outside repositionings 5 (0.8% of endocardial leads placed) Lead changes 3 (0.4% of total procedures) Infected IH explantations 0 (0.0% of all infected explantations) Infected outside explantations 21 (3.1% of total procedures) Noninfected procedures DC

=

dual chamber;

IH = in-house;

SC = single chamber

in-house implantations. The in-house infectious complication rates were O.O%, 0.52%, and 1.01% during the years 1986 to 1988, respectively. One patient in each of the years 1987and 1988 underwent evacuation of the generator pocket for hematoma evacuation 1to 3 days after the initial implantation. No subsequent infections developed in either of these patients.

Era 2 The total number of pacemaker procedures increased in the years 1990 to 1992 (Table 3), as did the number of new implants (see Fig 2). Likewise, the percentage of dualTable 3. Experience Between July 1989 and December 1992 (Era 2) Procedure

Number

Total procedures Total new implantations Transvenous SC Transvenous DC Epicardial SC Epicardial DC Battery changes Repositioned IH leads Periop repositionings Outside repositionings Lead changes Infected IH explantations Infected outside explanations Noninfected procedures DC

=

dual chamber;

875 709 (81.0% of total procedures) 308 (43.4% of new implantations) 382 (53.9% of new implantations) 11 (1.6% of new implantations) 8 (1.1%of new implantations) 79 (9.0% of total procedures) 36 (3.3% of endocardial leads placed) 32 (3.0% of endocardial leads placed) 13 (26.5% of leads repositioned) 12 (1.1%of endocardial leads placed) 5 (0.6% of total procedures)

IH

3 (37.5% of all infected explantations)

21 (2.4%of total procedures) =

in-house;

SC = single chamber.

592

FERGUSON ET AL SURGEONS AND CARDIAC PACING

Ann Thorac Surg 1994;5758%97

chamber implants as a percent of new implants increased, from 21.7% in the second half of 1989 to 79.0% in 1992 (see Fig 3). The total endocardial lead reposition procedures rate (expressed as a percent of new implants) increased during 1990 to 1992 as compared with 1986 to 1988 (Fig 4); the in-house lead reposition procedure rate during era 2 increased from 3.1% in 1990 and 1991 to 7.1%. To determine if this was a reflection of the overall increase in the number of leads implanted as a result of the increased number of dual-chamber implants, the in-house reposition procedure rate expressed as percentage of total number of endocardial leads implanted was examined, and was found to be 2.3%, 3.4%, and 4.1% during 1990 to 1992; of the 36 in-house patients who underwent lead repositioning during era 2, four were performed at an interval after implantation greater than 45 days. Explantation procedures for infections expressed as a percentage of total implants were 0.38%, 0.76%, and 1.5% during 1990-1992. The in-house implant infection rates during these years were 0.38%, 0.76%, and 0.76% respectively. One patient underwent evacuation of a hematoma during the second era. Since the subclavian introducer technique was used in 99% of implantations during the second era, the risk of pneumothorax was a potential complication during this time that was not present in era 1. The incidence of pneumothorax (expressed as a percent of implant procedures requiring cannulation of the vein) was 1.9%in 1990, 2.3% in 1991, and 2.7% in 1992; this includes 2 patients who had evidence of a very small pneumothorax on the postimplantation portable chest roentgenogram and were treated expectantly during this time, with resolution by the following morning.

Comparison of Era 1 and Era 2 The total number or procedures and the number of new implants increased significantly from era 1 to era 2 (see Figs 1,2). In contrast, the numbers of battery changes and noninfected procedures were similar (see Tables 2, 3).

ERA 1

z:

23

20 0

IH Ld Repos

IH lnfx

Hematoma

PTX

fl

Total Comr,

58

IH Ld Repos

IH lnfx

Hematoma

PTX

Total Comp

Fig 4. Comparision of complication rates between eras 2 and 2. (Hematoma = hematoma evacuation procedures; IH Ink = in-house infection explantation procedures; IH Ld Repos = in-house lead reposition procedures; PTX = pneumothorax; Total Comp = total complication rate [see text for further definitions].)

The distribution of single-chamber versus dualchamber implants was significantly different during era 2 as compared with all years in era 1(Table 4). During 1992, the implant criteria for W I or W I R pacemaker system placement was chronic atrial fibrillation with bradycardia, or supraventricular tachycardia that was uncontrolled medically in a patient who underwent atrioventricular nodal ablation. The infectious complication risks for in-house implants were not significantly different between the two eras (Fig 4), despite the increased number of procedures performed during the second era. The total lead reposition rate was increased in era 2 as compared with era 1 (Fig 5). However, when the in-house lead reposition procedure rates were expressed as a percentage of the total number of endocardial leads implanted the rates were not significantly different between

Table 4. Comparisons of Era 1 Versus Era 2 P

Variable

Value

Total no. of procedures (era 1 versus era 2) No. of new implantations (era 1 versus era 2) No. of endocardial leads placed Percent of dual-chamber placements Total complication rate (3.3% versus 6.6%) Frequency of IH lead repositioning Infectious complication rate Frequency of hematoma Frequency of pneumothorax

0.016"

a

Unpaired Student's t test.

IH

=

in-house.

,y2 analysis.

0.017"

ERA 1 E

8

2;

800 600 400

200

n

New Endo Lds

IH Ld

Repos

Tot Ld Repos

New Endo Lds

IH Ld Repos

Tot Ld Repos

0.003" 0.00Ob 0.006b 0.457b 0.713b 0.429b 0.003b

Fig 5. Comparison between total number of endocardial leads (Endo Lds) implanted and lead reposition (Ld Repos) rates between eras 1 and 2. (IH = in-house.)

FERGUSON ET AL SURGEONS AND CARDIAC PACING

Ann Thorac Surg 1994;57588-97

eras. The total number of endocardia1 leads implanted increased significantly between eras 1 and 2 as a result of the increased number of dual-chamber implants in era 2. There was a slight increase in the number of patients transferred from outside hospitals to Barnes Hospital for further treatment of a pacemaker problem during the second era, as manifest by the increased number of procedures on patients not categorized as in-house. Use of the subclavian technique for venous access resulted in an incidence of pneumothorax of 2.1% for era 2 (see Fig 4). Aggressive utilization of tube thoracostomy was made to minimize hospital stay following this complication. Because essentially all implantations were performed by cutdown on the cephalic or external jugular vein during era 1, comparison of the two techniques with regard to pnemothorax as a complication is not valid. The instance of ventricular perforation with tamponade and the number of hematoma complications were not different between eras. However, the total complication rates were significantly different (see Table 4), reflecting the incidence of pneumothorax in era 2 (see Fig 4).

Comment As health care reform is undertaken in the United States on a scale exceeding even the implementation of Medicare in 1965, bradypacing technology and practice is being closely scrutinized by government, third-party payors, and hospitals. Overall, bradypacing has come to be perceived as an example of the excessive application of technology that is thought by many to be partly responsible for the soaring increases in health care costs. This negative perception is supported by confusion on the part of outside observers with respect to the changes in manpower, technology, and costs that have occurred in the past decade [2].

Manpower The multiple-speciality involvement in cardiac pacemaker implantation is thought by some to lead directly to increased use of pacing therapy, and in some instances, to overuse to support the large number of physicians involved. Data from the most recently compiled Survey of Cardiac Pacing in the United States tend to bear out this perception [l].It was estimated from information from the six largest pacemaker suppliers that in 1989 a total of 89,445 new implantations and 21,055 generator changes were performed by 7,919 implanters from multiple medical disciplines at 3,400 centers in the United States. Of the 7,919 estimated implanters, however, only 571 (7%) responded to the survey, thus potentially limiting interpretation of the results; nevertheless, half of these respondents were surgeons, and it is from these data that the current estimate that approximately half the pacemaker implantations performed in this country are performed by surgeons is made. This percentage of surgical involvement in pacing has continued to decline throughout the past decade, however, when these results are compared with the results of previous surveys [l].Elimination of surgical involvement in pacing altogether might further streamline these manpower issues.

593

Technology Pacemaker industry data suggest that 35% to 40% of new implants placed in the United States are dual-chamber devices. Worldwide, this figure approaches only 30% [3]. This is at variance with the recommendations of a number of pacing groups, including the North American Society of Pacing and Electrophysiology (NASPE)’s task force on cardiac pacing [4] and the British Pacing and Electrophysiology Group (BPEG) [5]; their recommendations are that 70% to 80% of implants should be dual-chamber in accordance with the joint American Heart Association/ American College of Cardiology Guidelines for Pacing [6]. Indeed, the BPEG has stated that “the atrium should be paced unless contraindicated,” and Barold and Sanders [7] have characterized this attitude as the golden rule of pacing in the decade of the 1990s. The unwillingness of cardiothoracic surgeons to keep abreast of pacing technology has led some to argue against continued surgical involvement in pacing on this basis. Costs In theory, consolidation of pacemaker therapy delivery could control excesses, improve appropriate use of new and existing technology, and thereby limit costs; the highest quality of care ends up being the least expensive. This is the genesis of the so-called ”Centers of Excellence” concept. How this consolidation might occur, and what the driving force behind the consolidation (eg, government mandate versus assessment of quality and outcomes) of care might be, is unclear [l, 8, 91. For the most part, cardiac surgeons have yielded their involvement in cardiac pacing to other medical specialists because of (1)time considerations, (2) lack of interest in pacing technologic advances, (3) decreasing levels of reimbursement, (4) low levels of reimbursement as compared with other cardiac and thoracic procedures, and (5) outside pressure by medical colleagues or hospital administrators to do so, often against the wishes of the surgeons involved. This study was undertaken to critically assess the role of the cardiothoracic surgeon in the overall field of bradypacing at our institution. To the degree that these data are comparable with available data in the literature, they can be used to determine the overall effectiveness and quality of this current approach to pacemaker therapy that involves cardiothoracic surgery as an integral part of the program.

Era 1 The first era, 1986 to 1988, is thought to be representative of ”typical” cardiothoracic surgical involvement in pacing therapy as it evolved, and in some cases to be representative of pacing practices at a number of smaller practices around the country even today. This assumption is supported by data from the 1989 NASPE Survey [l];surgical implanters tended to work alone at implantation, to work in two hospital settings, and to work at larger institutions (median size, 330 beds) if they were ”high-volume” implanters (>12 implantdy). They tended to rely on industry representatives for intraoperative measurements, to work in an operating room, and to use active fixation leads in the atrium and passive fixation leads in

594

FERGUSON ET AL SURGEONS AND CARDIAC PACING

the ventricle. Overall, 68% of implants in 1989 were single-chamber devices with or without rate-response. In contrast to the typical experience in the survey, however, vascular access during era 1 was accomplished by the cephalic vein cutdown technique. The concerns regarding follow-up during era 1 were also borne out by the Survey results; in terms of followup, a pacemaker service existed in only 43% of the respondent’s primary institution, whereas 68% of respondents used telephone monitoring for follow-up with or without clinic visits. Moreover, 19% of all pulse generators implanted were not reprogrammed from factory settings either at operation or during the follow-up interval.

Era 2 The second era, 1990 to 1992, reflects a number of changes that were made in cardiac pacing at Barnes Hospital. To address the issues of follow-up, a formal Pacemaker Service was established. Generator implantations continued to be performed in the operating room by a cardiothoracic surgeon with specific interest in cardiac pacing and surgical electrophysiologic problems. The subclavian approach was used for venous access in almost all implantations, and this facilitated placement of dual-chamber devices when medically indicated. The ability to provide adequate telephone monitoring as well as clinic follow-up for all patients further facilitated appropriate evaluation and reprogramming of dual-chamber and rate-responsive devices. These three factors-simpler access, adequate follow-up for more complex devices, and the general support provided by a Pacemaker Service-were primarily thought to be responsible for the increase in new implant volume in era 2 (see Fig 2). Clearly, the necessity for adequate follow-up has become more important in the setting of the advanced technology involved in dualchamber pacing and rate-responsive pacing to justify the medical benefit and increased costs of this technology; in contrast to the results of the NASPE 1989 Survey, virtually all patients in the Washington University Pacemaker Clinic have undergone some type of reprogramming of their device at the time of the initial clinic follow-up visit. Currently, 1,450 patients have been followed up by telephone monitoring, and the Clinic performs 380 outpatient and 250 inpatient evaluations per year. By 1992, the percent of dual-chamber implants was similar to that recommended by NASPE and the BEPG [4, 51. The reasons for this include (1) familiarity with the physiologic and technologic aspects of dual-chamber pacing, (2) ease of implantation, and (3) provision of an adequate follow-up mechanism. Currently, the only patients undergoing implantation of a VVI or VVIR device have intractable atrial fibrillation and bradycardia, or have undergone atrioventricular node ablation for a medically refractory supraventricular arrhythmia.

Comparison of Era 1 and Era 2 These changes in the practice of pacing between eras 1 and 2 resulted in an increase in the overall complication rates in era 2 (see Table 4, Fig 5 ) , as a result of lead reposition and pneumothorax complications. However, when the lead reposition rate was analyzed based on the

total number of endocardia1leads implanted, the two eras are not significantly different, as demonstrated in Figure 4. Furthermore, the overall difference in complication rates between eras 1 and 2 is essentially accounted for by the presence of pneumothorax as a complication in era 2; the incidences of other types of complications were not different between the two eras. The overall complication rates for pacemaker implantation has been often quoted to be between 5% and 20%, with infectious complication rates between 3% and 19% [lo-121. Three more recently published studies have documented lower complication and infection rates [13--15]. Parsonnet and associates [ 151 analyzed the pacemakerimplantation complication rates at a single institution where implantations were performed by 29 implanting physicians, both surgeons and nonsurgeons. The overall complication rate was 5.7%, but late complications were not included in the analysis. The percent of DDD devices implanted was approximately 55%. They found a substantially large incidence of complications experienced by implanters who performed fewer than 12 implantations per year; this number of 12 is of interest, because the “average” number of new implantations per implanter during 1989 according to the NASPE survey was 11 (89,445 implants/7,919 implanters) and total procedures was 13.9 (110,500 implants/7,919 implanters) [l].The Parsonnet study also demonstrated increased complication rates associated with low-volume (<12/y) nonsurgeon implant procedures as compared with other procedures, including low-volume surgeon implanters. In addition, physicians regardless of specialty who paid little attention to surgical technique had a higher associated complication rate than did those who paid considerable attention to technique. The remaining large series reported in the literature have been from institutions where surgeons performed the majority of implantations in operating room environments [13, 141. Brodman and colleagues [13] reported a 0.56% infection rate for 7,435 generator implantations in 4,333 patients over a 17-year period, similar to the overall results of the present study; a subset of these patients undergoing DDD device implantations had a 0.4% infectious complication rate, again similar to the era 2 results. Unfortunately, because of the multispeciality nature of pacing implant therapy, accurate recent data in terms of complication rates from implantations performed by nonsurgeons are not readily available for comparison with these published series or with the present study. Although the low infectious complication rate in the present study might be attributable to the operating room implantation environment, data from the analysis were not available to support or refute this point. In the NASPE 1989 Survey, the majority of implantations reported were performed in the operating room (66%)as compared with a catheterization facility (23%) [l]. Two reports have described comparison of operating room versus catheterization laboratory implantation experiences in small numbers of patients [16, 171. In Miller and associates’ [16] series the implantations were all performed by a cardiothoracic surgeon, including the ones in the catheterization laboratory. Modest cost savings were demon-

Ann Thorac Surg 1994;57588-97

strated in both studies in small, selected groups of patients. Hess and co-workers [18]reported a series of 171 patients undergoing implantation in the catheterization laboratory with no infectious complications. Surgical experience, expertise, and knowledge of technique appear to be factors of greater importance than locale in reducing pacemaker implantation infectious complications to a minimum. From a practical point of view, however, the ability to stagger pacemaker implantations among six other cardiac and thoracic operating rooms at Barnes Hospital greatly facilitates the efficiency of implantation and the efficient use of the operating room space and personnel over potential device implantation in two dedicated electrophysiology laboratories. This increase in efficiency appears to more than offset the increased cost of the operating room environment. What would be the potential advantages of taking surgeons "out of the loop" for pacing device implantation? The first involves increasing the "efficiency" of care because the implanting physician would in most cases be providing the follow-up care as well. However, removing surgeons from the pacing loop would result in the loss of second-opinion input to the clinical situation, assuming the implanting surgeon had a working knowledge of the basic principles of dual chamber and rate-responsive pacing. This has been alluded to when discussions regarding self-referral involving cardiac pacing arise [19, 201, and the potential deleteriousness of this self-referral situation has recently been demonstrated with respect to coronary artery disease treatment by angioplasty [21]. In contrast, the data presented here argue strongly for establishment of a multispeciality Pacemaker Service staffed by individuals with demonstrated interest and expertise in pacing therapy. The second "advantage" in taking surgeons out of the loop is that it would permit catheterization laboratory implantations with associated lower implantation costs. As mentioned above, there can be other efficiency considerations to this decision. In addition, whether implantations performed in the catheterization laboratory rather than an operating room facility decreases the overall cost of implantation must take account the cost of treatment of patients in whom infections of the systems develop [22]. In the present study, a total of 11 patients underwent explantation of an infected pacemaker system; 3 of these were transferred from an outside hospital. Seven of these 11 patients underwent explantation of infected systems between 1988 and 1992, the years for which complete financial data were available. These infectious complications resulted in an average additional length of stay of 15.6 9.9 days (range, 9 to 31 days); the average fully loaded hospital cost was $26,886 k 18,713 (range, $4,466 to $54,525). Four patients had private insurance and 3 were covered by Medicare. The overall average net hospital income was $2,091 5 $12,349 (range, $22,616 to -$14,589); for commercial insurers this figure was $9,986 2 $11,819, but for Medicare patients this care resulted in an average net loss of $5,804 k $7,066 per patient. For patients with infected pacemaker systems covered by Medicare, the average losstinpatient day was $504.69 per day. The third "advantage" in taking surgeons out of pacing

*

FERGUSON ET AL SURGEONS AND CARDIAC PACING

595

therapy is to circumvent of the lack of pacing technology expertise on the part of surgeons performing the implantations. This issue of inadequate expertise by surgeons in the area of cardiac pacing should not be taken lightly. To date, the American Board of Thoracic Surgery remains the only specialty board that requires a minimum number of pacemaker implantations to sit for the certifying examination. The American Heart Association/American College of Cardiology joint guidelines for the indication for pacemaker therapy [6] only suggest that the implanting physician have experience with pacemaker technology, implantation techniques, and follow-up. Of note, NASPE is soon to publish their selected Training Requirements for Permanent Pacemaker Selection, Implantation, and Follow-up; these guidelines specifically address the multispeciality nature of these procedures. Despite the singular nature (to date) of this requirement by the American Board of Thoracic Surgery, however, the degree of exposure to pacing required by the American Board of Thoracic Surgery has not changed over the past decade, during a period of time where dramatic advances in pacing technology have occurred. With the small amount of emphasis placed on cardiac pacing during most residency programs, however, it is unlikely that adequate exposure to the more subtle aspects of dual-chamber, rate-responsive bradypacing is being acquired in this setting without a concerted effort on the part of the resident and faculty. The establishment of a pacemaker service consisting of input from cardiologists, surgeons and electrophysiologists (if available) has several advantages as demonstrated in this study: (I) eliminate the problems associated with self-referral and single-physician provider from diagnosis through therapy; (2) maximize the effectiveness of the multispecialty involvement in pacing that exists at many institutions; and (3)provide for complete follow-up of this expensive technology, both in terms of patients welfare and in terms of issues of quality of care. Furthermore, this study demonstrates that (1) active cardiothoracic surgical involvement in cardiac pacing can be an integral component in a "state of the art" approach to pacing; (2) expertise in "state of the art" pacing technology can be readily acquired by cardiothoracic surgeons interested in pacing, thus eliminating the need for an adjunctive nonsurgeon's presence at the implantation; (3) that surgical involvement in cardiac pacing contributes to minimization of implantation-related complications; and the expenses involved in treating those complications; (4) that additional training of cardiothoracic residents in cardiac pacing principles and techniques will be necessary above what has been done in the past to allow them to keep abreast of pacing technology and therapy; and (5) that the advantages of surgical involvement in pacing that were paramount in the beginning of pacing (infection rates, complication rates, ability to handle the complications induced by a specific procedure) are still present in 1993. As health care reform moves forward, these issues will become more and more important. The available data suggest that to control (cf "regulate") costs, large volume centers performed by individuals doing a large volume of procedures yields the best technical results overall, re-

596

FERGUSON ET AL SURGEONS AND CARDIAC PACING

gardless of t h e therapy. As h a s been demonstrated by other centers [13], embracing the concept of a n integrated Pacemaker Service provides for the best overall quality of care, a n d can be thought of as a n intermediate step toward t h e Pacemaker ”Center of Excellence.” Continued cardiothoracic surgical involvement i n bradypacing is a logical and, as demonstrated i n this study, highly desirable component of this concept.

Ann Thorac Surg 1994;5758%97

11. 12. 13.

References 1. Bernstein AD, Parsonnet V. Survey of cardiac pacing in the United States in 1989. Am J Cardiol 1992;69:331-8. 2. Brinker JA. Who is the patient’s physician [Editorial]? Intell Rep Cardiac Pacing Electrophysiol 1989;8(2):14. 3. Goldman BS, Fraser JD, Morgan CD. Survey of cardiac pacing in Canada. Can J Cardiol 1991;7391-8. 4. Parsonnet V, Escher DJ, Furman S, et al. Indications for dural-chamber pacing. PACE 1984;7:318-9. 5. Clarke M, Sutton R, Ward D, et al. Recommendations for a pacemaker prescription for symptomatic bradycardia. Report of a working party of the British Pacing and Electrophysiology Group. Br Heart J 1991;66:185-91. 6. Dreifus LS, Fisch C, Griffin JC, Gillette PC, Mason JW, Parsonnet V. Guidelines for implantation of cardiac pacemakers and antiarrhythmic devices. J Am Coll Cardiol 1991; 18:l-13. 7. Barold SS, Sanders RS. Rate-adaptive cardiac pacing: cost versus technology versus patient benefit. Am Heart J 1993; 126:182a34. 8. De Belder MA, Linker NJ, Jones S, Camm AJ, Ward DE. Cost implication of the British Pacing and Electrophysiology Group’s recommendations for pacing. Br Med J 1992;305: 861-5. 9. Ray SG, Griffith MJ, Jamieson S, Bexton RS, Gold RG. Impact of the recommendations of the British Pacing and Electrophysiology Group on pacemaker prescription and on the immediate costs of pacing in the Northern Region. Br Heart J 1992;68:531-4. 10. Ruiter JH, Degener JE, Van Mechelen R, Bos R. Late purulent

14. 15. 16.

17.

18. 19. 20. 21. 22.

pacemaker pocket infection caused by Staphylococcus epidermidis: serious complications of in situ management. PACE 1985;8:903-7. Jara FM, Toledo-Pereyra L, Lewis JW, Magilligan DJ. The infected pacemaker pocket. J Thorac Cardiovasc Surg 1979; 78:29a300. Holmes DR, Hayes DL. Pacemaker implantation techniques. In: Saksena S, Goldschlager N, eds. Electrical therapy for cardiac arrhythmias. Philadelphia: Saunders 1990:173-91. Brodman R, Frame R, Andrews C, Furman S. Removal of infected transvenous leads requiring cardiopulmonary bypass or inflow occlusion. J Thorac Cardiovasc Surg 1992;103: 649-54. Gross JN, Moser S, Benedek ZM, Andrews C, Furman S. DDD pacing mode survival in patients with a dual-chamber pacemaker. J Am Coll Cardiol 1992;19:1536-41. Parsonnet V, Bernstein AD, Lindsay BD. Pacemakerimplantation complication rates: an analysis of some contributing factors. J Am Coll Cardiol 1989;13:917-21. Miller GB, Leman RB, Kratz JM, Gillette PC. Comparison of lead dislodgement and pocket infection rates after pacemaker implantation in the operating room versis the catheterization laboratory. Am Heart J 1988;115:1048-51. Stamato NJ, OToole MF, Enger EL. Permanent pacemaker implantation in the cardiac catheterization laboratory versus the operating room: an analysis of hospital charges and complications. PACE 1992;15:223&9. Hess DS, Gertz EW, Morady F, Scheinman M, Sudduth BK. Permanent pacemaker implantation in the cardiac catheterization laboratory. Cathet Cardiovasc Diagn 1982;8:453. Falk RH. Impact of prospective peer review on pacemaker implantation rates in Massachusetts. J Am Coll Cardiol 1990;15:1087-92. Griffith MJ, Bexton RS, McComb JM. Financial audit of antitachycardia pacing for the control of recurrent supraventricular tachycardia. Br Heart J 1993;69:272-5. Topol EJ, Ellis SG, Cosgrove DM, et al. Analysis of coronary angioplasty practice in the United States with an insuranceclaims data base. Circulation 1993;87:1489-97. Byrd CL, Schwartz SJ, Hedin N. Cardiac pacing. Lead extraction: indications and techniques. In: Cardiology clinics. Philadelphia: Saunders, 1992:73548.

DISCUSSION DR MICHAEL D. HOROWITZ (Pascagoula, MS): I very much enjoyed hearing this presentation and I appreciate the opportunity of reviewing the manuscript in advance. In a broad sense this report addresses an issue that affects us all as thoracic surgeons. The issue is who should be doing surgical procedures in the chest? This question is applicable not only to pacing but to other procedures, such as bronchoscopy and thoracoscopy. In this study Dr Ferguson and associates demonstrate excellent results with a team approach wherein the pacemaker implantation was performed by a thoracic surgeon and then the follow-up was performed by appropriately trained and interested cardiologists. Currently the American Board of Thoracic Surgery is the only board that requires experience implanting pacemakers for board certification. Nevertheless, half of all pacemaker implantations in America are performed by physicians who are not ABTS eligible. In the community where I currently practice, a substantial number of pacemakers are implanted by general surgeons. This long-term arrangement is the result of the absence of a thoracic surgeon until very recently. I would like to ask Dr Ferguson if he has any information or data, or perhaps any thoughts, on the role of general surgeons in pacemaker implantation.

DR KAMAL A. MANSOUR (Atlanta, GA): I thank Dr Ferguson for sending me the manuscript for review. This is a timely and excellent report that sets the ideal for pacemaker implantation, that is, a combined medical-surgical approach. It stresses the need for having cardiothoracic surgeons interested in pacing and implantation techniques. Until a few months ago, Dr Joe Miller and I have implanted all pacemakers at Emory, more than 5,000 units. However, with the establishment of electrophysiologic laboratories across the country, the pendulum has shifted almost totally to cardiologists and electrophysiologists for pacemaker implantation in their laboratories, whether electrophysiologic or catheterization laboratories. We should look really for reports of large series of pacemakers documenting the infection rate and the rate of other complications such as ventricular perforation, pneumothorax, and hemothorax in patients undergoing pacemaker implantation in the electrophysiologic laboratory versus those having implantation in the operating room by surgeons. There still remains an important place for epicardial pacemaker implantations by thoracic surgeons. Epicardial pacing should be the primary approach in the newborn and during open heart operations. It should also be employed whenever endocardia1 pacemaker fails, and in patients with endocarditis of the tricuspid

Ann Thorac Surg 1994:57:58%97

valve or a prosthetic tricuspid valve or right ventricular apical electrical silence. Anatomotic anomalies of the right atrium and right ventricle may also dictate epicardial implantation. I certainly agree with Dr Ferguson’s conclusions that active surgical involvement in cardiac pacing by cardiothoracic surgeons can be an integral component in ”a state of the art” approach to pacing and that cardiothoracic residents should be conversant in cardiac pacing principles and techniques. My only question to Dr Ferguson is, how were you able to transform a realistic situation such as ours to an idealistic situation such as yours? I certainly enjoyed the presentation very much. DR JOHN M. KRATZ (Charleston, SC): I enjoyed your report also. Your setup in St. Louis is very much like ours in Charleston with an electrophysiologic cardiologist following and a surgeon always doing the implantation. Addressing the issue of catheterization laboratory versus operating room, when we first started as surgeons putting in the pacemakers, we did them all in the catheterization laboratory; we got tired of being second-rate citizens and getting to do the procedures at 8 o’clock at night when they finished, and we moved down to the operating room for that reason only. We studied the differences between the two results in the two places, and it came out exactly the same; incidence of infections was very low, less than 1% in both places, and complications were the same. Our pediatric electrophysiolic cardiologist continues to do these procedures in the catheterization laboratory with good results; we continue to do them in adults in the operating room with good results. So I think that who does it and their training and their interest is very important, but probably catheterization laboratory versus operating room is not an important question. DR WILLIAM A. COOK (North Andover, MA): I would like to present the view-point of the privately practicing thoracic surgeon to this discussion. One thing that is very important is the enormous control the cardiologist has over this situation. If he or she does not refer the case to the local surgeon, the surgeon does not have any input into how the procedure gets done. Also, as it is looked upon as a fairly simple procedure, a lot of this work is not being done in tertiary institutions. It is being done in smaller hospitals all across America. To give you the flavor of how this works, in one hospital where I work I started pacemaker implantation and did them all for 5 or 6 years. Then the cardiologists started hearing that cardiologists were becoming involved in pacemaker implantation, and that fees were available for their being there. This was the critical issue. Suddenly the cardiologists were coming to me and saying, “I want to be involved in pacemaker implantation.” I asked them, ”For what purpose?” That was the wrong question to ask, because in that hospital where I started the whole business of pacemaker implantation, I no longer do any pacemaker implantations. They turned to my general surgical colleagues and said, ”Look, you make the incision; we will manipulate the catheter.” I assure you that is probably the most common arrangement now in this country. The other part of this is that some of the younger doctors are coming out of cardiology training programs fully convinced that they can d o the whole procedure from skin to skin without any surgeon involved at all. They can use their leverage in the community setting to force the thoracic surgeon to help them establish their credentials for doing the procedure. In one such situation I helped a new cardiologist do some cases. The creden-

FERGUSON ET AL SURGEONS AND CARDIAC PACING

597

tials committee asked me what I thought about his potential for doing the procedure. I said I was sure that he knew everything he needed to know about the cardiologic part of it but I did not think he was a very facile surgeon and that he could get into trouble. I suggested that, at a minimum, he should have a thoracic surgeon backing him up, in other words, available, any time he did one of these procedures, and that he have a clear understanding that that was his responsibility. About 4 weeks later I was doing a case of my own at a hospital across town. I received a telephone call that that cardiologist had a patient with an artery bleeding like crazy. I had to leave my patient and go across town to take care of his patient. This was done in open defiance of the conditions of his credentialing, but it did not change his pattern of practice. DR FERGUSON: I would like to thank all of the discussants for their comments. In response to Dr Horowitz‘s question, no data are available in terms of large surveys regarding what percentage of the surgical implantations done in t h s country are done by general surgeons and what percent are done by cardiothoracic surgeons. I would agree completely with Dr Cook that in the typical situation, the majority of surgeons are involved in pacing in conjunction with a cardiologist, and fees are being collected by both individuals for that implantation; and as we all know, that will be no longer possible very soon. That will even lessen the incentive of cardiologists to involve surgeons in the practice of cardiac pacing. In response to several other comments Dr Cook made, there are good data to show that individuals who implant fewer than 12 pacemakers a year have significantly higher pacemaker complication rates than those who implant more than 12 a year. There are also good data to show that pacemaker implantations done in small hospitals have a higher complication rate. One of the purposes of putting this information together was because the government is going to be mandating who can and who cannot put pacemakers in, not so much in terms of qualifications but in terms of who gets payment. And since we obviously have a vested interest in pacing at Barnes Hospital, we do not want somebody coming and telling us that they are not reimbursing surgeons for Medicare patients undergoing pacemaker implantation; 85% of the implants done in this country are in Medicare patients. Unfortunately, there are no good data in the cardiology or surgical literature describing clinical results from a large implant series performed by nonsurgical implanters. There are some smaller anecdotal series and there are good data, as was indicated by Dr Kratz, regarding catheterization laboratory versus operating room implantation. I would agree with him completely: it is not where the implantation is done, it is who is doing it and what their training and expertise is. We as cardiothoracic surgeons, both cardiac and thoracic, have turned over a large amount of our practice to other specialties over the last 25 years, and with circumstances such as thoracoscopic surgery, implantation of implantable cardioverterdefibrillators by nonthoracotomy techniques, and so forth, there is a real possibility that pressures to give up areas of wellestablished expertise by cardiothoracic surgeons in a number of areas will continue. This report was designed as an attempt to put some data out there that demonstrate current surgical practice in the setting of a combined medicalisurgical approach to cardiac pacing in 1993. Hopefully information such as this will be used in the future decisions that are made.