- Email: [email protected]
Reliability, cost-effectiveness, and safety of reuse of ancillary devices for ERCP
Reliability, cost-effectiveness, and safety of reuse of ancillary devices for ERCP Fre´de´ric Prat, MD, PhD, Jean-Franc¸ois Spieler, Msc, Sabine Paci, PharmD, Coralie Pallier, PharmD, Jacques Fritsch, MD, Andre´ Daniel Choury, MD, Gilles Pelletier, MD, Sylvie Raspaud, PharmD, Patrice Nordmann, MD, PhD, Catherine Buffet, MD Le Kremlin-Biceˆtre, France
Background: The choice between reusable and single-use devices for ERCP depends on various medical and economic criteria. This study evaluated the reliability and the safety (risk of crosscontamination) of reusable devices. A cost analysis of the use of reusable devices also was conducted. Methods: All patients referred for ERCP that required use of a sphincterotome or a retrieval basket were eligible for inclusion in a clinical study of 4 different devices (3 types of sphincterotome, 1 type of retrieval basket). All devices were steam sterilized. Before each use, each device was subjected to bacteriologic and virologic tests (hepatitis C virus, hepatitis B virus markers). Devices were examined before and after each procedure. The numbers of safe and efficient procedures that could be performed with each device were assessed. Three strategies were compared in a cost analysis: internal reprocessing (strategy 1), external reprocessing (strategy 2), and single-use (strategy 3). Inputs used were the results of the clinical study, hospital data for 1 year of endoscopic activity, and market prices. Results: A total of 342 patients underwent the following procedures: sphincterotomy (248 patients), stent insertion (59 patients), use of basket without sphincterotomy (14 patients), and diagnostic ERCP/unsuccessful cannulation (21 patients). At the time of ERCP, 36 patients had viral or bacterial infection. Fifty instruments were used (20 single-lumen sphincterotomes, 10 double lumen sphincterotomes, 20 retrieval baskets). Overall, the median number of efficient uses per device was 10. The median number of efficient uses by each type of device was the following: single-lumen sphincterotome, 12; double-lumen sphincterotome, 8; and, retrieval baskets, 10. All virologic and bacteriologic tests for all instruments were negative. The cost-optimization analysis found that strategy 1 is cost effective (e37,283/y) compared with strategy 2 (e40,101/y) and especially with Strategy 3 (e115,210/y). Conclusions: Reuse of the sphincterotomes and baskets evaluated in this study during ERCP is safe in terms of infectious hazards. Because they endure numerous uses, reusable instruments are cost effective, especially when compared with single-use accessories. (Gastrointest Endosc 2004;60:246-52.)
Interventional endoscopy contends with two conflicting requirements: optimal patient safety, by elimination of infectious risks and procedure-related complications, and control of costs in an era of decliningreimbursement,expensiveadvancesintechnology, and growing demands for health care by an aging population. Ancillary devices used in intervenReceived December 12, 2003. For revision January 12, 2004. Accepted February 24, 2004. Current affiliations: Service d’he´patogastroente´rologie, Inserm U537, Ste´rilisation centrale, Service de Microbiologie, CHU Biceˆtre, Le Kremlin-Biceˆtre, France. This work was presented in part at Digestive Diseases Week, May 20-23, 2001, Atlanta, Georgia (Gastrointest Endosc 2001;53:AB76). Reprint requests: Dr. Fre´de´ric PRAT, Service d’he´patogastroente´rologie, CHU Biceˆtre, 78 rue du Ge´ne´ral Leclerc, 94275 Le Kremlin-Biceˆtre, France. Copyright Ó 2004 by the American Society for Gastrointestinal Endoscopy 0016-5107/$30.00 PII: S0016-5107(04)01685-2 246
GASTROINTESTINAL ENDOSCOPY
tional endoscopy are representative of this dilemma: single-use devices supposedly are devoid of technical flaws and infectious agents, but economic pressure may dictate the choice for reusable devices.1-4 Market prices set by device manufacturers and regulations established by health authorities often are deciding factors in this choice. However, data from scientific studies may help to achieve an optimal balance between patient safety and cost containment. It was our hypothesis that single-use devices were, by definition, optimal and that only reusable devices required assessment in terms of safety, especially transmission of infectious agents and technical reliability. Therefore, a 2-part study was conducted: a prospective clinical evaluation of standard reusable ERCP accessories (infectious safety and reliability) and a comparative cost analysis of single-use and reusable devices by using the results of the clinical evaluation and accounting data from our hospital. VOLUME 60, NO. 2, 2004
ERCP ancillary devices: reliability, cost-effectiveness, and safety of reuse
PATIENTS AND METHODS Safety and reliability of ERCP devices Patients. Patients referred for interventional biliary endoscopy were eligible for inclusion in the study if use of a sphincterotome and/or a retrieval basket were anticipated. Procedures included were the following: endoscopic sphincterotomy, regardless of indication; stone extraction without sphincterotomy; stone extraction with a previous sphincterotomy; and diagnostic ERCP, requiring use of a sphincterotome for selective cannulation. Exclusion criteria were the following: diagnostic ERCP by using a standard catheter, therapeutic ERCP in which a single-use device was required, and a non-biliary or a pancreatic therapeutic procedure. The study was approved by our local ethical review board. Devices. Four distinct types of device were studied: two different single-lumen sphincterotomes, with a monofilament, 3-cm-long cutting wire and a short cannulation tip (KD-5q and KD-21q; Olympus Optical Co., Ltd. Tokyo, Japan); a double-lumen sphincterotome (KD-6G-12q1; Olympus); and, a medium-sized braided wire retrieval basket (FG-22q; Olympus). Hypotheses. The primary study end point was the number of times the ERCP accessory device could be used efficiently and safely. Based on a hypothesis that each type of sphincterotome could be used a mean of 7 6 2 times,5-7 the assessment of 10 items of each type (i.e., 30 sphincterotomes overall) appeared to be adequate (95% CI[5.76, 8.24]). The number of uses for retrieval baskets was a secondary end point. Two different devices (sphincterotome and retrieval basket) usually are used during most procedures, but sphincterotomes are used, overall, 1.5 times more often than retrieval baskets. Thus, the number of baskets to be studied was set at 20. A sphincterotome is the first device used in nearly 98% of the therapeutic ERCPs performed per year in our department (approximately 500 ERCPs/y). It was calculated that slightly more than 200 patients had to be enrolled in the study (30 sphincterotomes 3 7 uses = 210 patients). However, it was decided to continue the enrollment of patients until all the devices had been discarded.
Device reprocessing and microbiologic studies Reprocessing. Before the first use, each device was steam sterilized in an autoclave. The devices then were reprocessed according to guidelines established by the European Society of Gastrointestinal Endoscopy, which include 10 steps: immediate immersion and washing in fresh detergent solution after the endoscopic procedure, dismantling and flushing with detergent, brushing with a cleaning brush, US cleaning (20 minutes at 608C), rinsing and flushing under running clean water, drying, sterile packaging, transfer to the central sterilization unit of the hospital, steam sterilization (1348C for 18 minutes), and return from the sterilization unit to the endoscopy unit for storage.8 The sphincterotome handles were packed separately and steam sterilized in the same manner as sphincterotomes and baskets. VOLUME 60, NO. 2, 2004
F Prat, J-F Spieler, S Paci, et al.
Microbiologic studies. Each sterilized device was tested before use by flushing the injection port under sterile conditions with 2 mL of sterile saline solution. Samples were collected for bacteriologic and virologic studies. Briefly, one sample was used for culture in 4 different media (aerobic, anaerobic, Lowenstein’s medium for mycobacteria, and Sabouraud’s medium for yeasts and fungi), as well as specific media for Helicobacter pylori and Campylobacter species. Another sample was tested for hepatitis C virus (HCV) by quantitative polymerase chain reaction (PCR) amplification and hepatitis B virus (HBV) markers (HBs Ag and HBe Ag). The detection threshold for this HCV PCR test (Amplicor 2.0; Roche, Basel, Switzerland) is 100 copies. Patients also underwent serologic testing for HBV and HCV (HBs Ag, anti-HBs Ab for HBV, total anti-HCV immunoglobulins for HCV), unless their viral status was known. For patients who tested positive for serum viral markers, reverse transcriptase–PCR for HCV and serum HBV b-DNA also were obtained. With an expected frequency of iatrogenic infection of 0.1%: 95% CI[0.001, 0.005], 153 samples were sufficient to detect cross-contamination (data from epi2002 software). Thus, virologic and bacteriologic tests were discontinued after inclusion of 200 patients. However, this calculation reflected a broad variety of infections, but it did not take into account the number of patients actually at risk for each specific type, especially for viral species. Assessment of devices A double-tracking system was used to monitor all devices through successive ERCPs and reprocessings: each device was given an identification number, which was marked on the device with indelible ink. Records of each use, including date and patient identification, were kept in a registry and separately by using an individual form for each patient. Also recorded on the form were the diagnosis, the procedure performed, infectious risk factors (uncontrolled septicemia, positive blood cultures, positive viral markers, etc.), and the observations about the device. Each device used during a procedure was carefully examined and tested by the endoscopist immediately upon completion of the procedure. The device was graded as suitable or unsuitable for another procedure. To be judged suitable, all devices had to be fully operable. Sphincterotomes had to have a straight cutting wire, patent lumen(s), no leakage when flushed, effective bending at the tip, and an intact electrical connection. Dormia baskets had to have widely unfolding wires and smooth motion of the basket into the sheath when retracted. An instrument that did not fulfill all criteria, or if any part of it was broken, was considered unsuitable for further use and was immediately discarded. To standardize assessment among endoscopists, all devices used throughout the study were assessed by one of 3 senior endoscopists (F.P., J.F., A.D.C.). All endoscopists were blinded as to the number of previous reprocessings of each of the devices used throughout the study. The mean (standard deviation) and median numbers of uses were calculated per type of device as well as the range. GASTROINTESTINAL ENDOSCOPY
247
F Prat, J-F Spieler, S Paci, et al.
ERCP ancillary devices: reliability, cost-effectiveness, and safety of reuse
COST ANALYSIS The 1-year cost of 3 strategies for controlling costs of ERCP accessories was assessed: strategy 1 (that used in our hospital) consisted of in-house reprocessing of accessories; strategy 2 consisted of outsourcing of accessory device reprocessing to a third party; strategy 3 consisted in using only single-use devices. Each strategy generates different types of costs: for strategy 3, costs pertain essentially to purchase of the accessory devices, whereas, strategies 1 and 2 include purchase and reprocessing costs. The global cost of strategy 1 was estimated by using accounting data from our center (a 997-bed university-affiliated hospital in the greater Paris area). The overall costs for the other two strategies were determined from the number of procedures performed in our center, the market prices for the accessories used, and reprocessing costs. The comparison was based on the use of double-lumen sphincterotomes and retrieval baskets. Single-lumen sphincterotomes were not analyzed because nearly all disposable sphincterotomes currently available have two lumens. This study was approved by our local ethical review board. For purposes of the cost analysis, the median number of uses per reusable device as determined in the clinical study was taken as the number of uses per reusable instrument: this was 8 for reusable double-lumen sphincterotomes and 10 for reusable retrieval baskets. For the number of endoscopic procedures preformed, the actual number of ERCPs performed in our unit during 1 year (2002) was used. Purchase costs were the actual costs negotiated between the hospital and the manufacturers, for both reusable and single-use devices. Reprocessing costs Strategy 1 (in-house reprocessing). Reprocessing costs were estimated by using accounting data from the hospital. Sterilization takes place in a central sterilization unit, a division of the pharmacy department. The following costs were identified directly from hospital accounting records: personnel charges and wages, consumables (sterilization products, disposable items, linen, etc.), and purchase and maintenance of reprocessing equipment. Personnel costs are divided into two categories: direct costs (actual time spent working in the sterilization unit) and indirect costs (education, etc.). The sterilization unit staff includes 20 full-time nurse’s aides, 7 full-time nurses, one full-time manager, and a half-time pharmacist. Expenditures for utilities (water, electric power) are charged globally to the hospital; the relative proportion of these expenditures for the sterilization unit was calculated by using the fraction of the surface area of the hospital occupied by the sterilization unit. The cost of sterilization for endoscopic accessories was estimated by calculating a relative workload ratio for endoscopic accessories reprocessing vs. the total workload of the sterilization unit and then by applying this ratio to the costs of consumables, reprocessing equipment, and indirect personnel costs, and also to direct costs for personnel. Strategy 2 (external reprocessing). An estimate of the costs for externalized sterilization of endoscopic 248
GASTROINTESTINAL ENDOSCOPY
accessories was based on mean market prices (in euros, based on prices quoted in 2002). The cost of waste disposal was not taken into account. Detailed data used for calculations for each of the 3 strategies are presented in Table 1.
RESULTS Patients and procedures performed A total of 342 consecutive patients (144 men, 198 women; mean age 69.7 [17.2] years) were enrolled in the study. The indication for ERCP was the following: suspected choledocholithiasis, 235 patients (68.7%); cholestasis with suspected biliary stricture, 79 patients (23.1%), 12 of which were benign; chronic pancreatitis, 10 patients (2.9%); biliary leak, 8 patients (2.3%); and suspected sphincter of Oddi dysfunction, 10 patients (2.9%). Thirty-six (18%) of the 200 patients for whom virologic and bacteriologic testing was obtained were identified as potential sources for transmission of infection (cross-contamination). These patients had the following infections: chronic HCV infection (n = 11), chronic HBV infection (n = 4), HIV infection (n = 5), and septicemia with positive blood cultures (all gram-negative strains) at the time of ERCP (n = 16). Procedures performed were the following: endoscopic biliary sphincterotomy with stone extraction, 194 patients (56.7%); biliary sphincterotomy without stone extraction, 49 patients (14.3%); sphincterotomy with insertion of a biliary stent, 55 patients (16.1%); stent exchange, 4 patients (1.2%); stone extraction (previous biliary sphincterotomy), 14 patients (4.1%); pancreatic sphincterotomy, 5 patients (1.5%); and non-therapeutic ERCP (i.e., intended therapy but ductal opacification alone), 15 patients (4.4%). ERCP was unsuccessful in 6 patients (1.7%), but these procedures were included in the analysis because the study devices were used. There was no case of successful cannulation followed by unsuccessful sphincterotomy. Immediate complications were observed in 3 patients: two post-sphincterotomy hemorrhage, one duodenal perforation. Failures and complications occurred with devices in their first to 13th use. There was no correlation between failure and complications and dysfunction of the devices. Microbiologic analyses Analyses for bacterial and yeast contaminants were negative except for 4 devices. The organisms identified were coagulase-negative Staphylococcus (3/4) and hemolytic Streptococcus species (1/4). Because these organisms are not usually digestive-tract VOLUME 60, NO. 2, 2004
ERCP ancillary devices: reliability, cost-effectiveness, and safety of reuse
F Prat, J-F Spieler, S Paci, et al.
Table 1. Elements of cost calculation (costs for 1 y of endoscopic activity) A) Strategy 1: in-house reprocessing Reusable double-lumen sphincterotome
Retrieval basket
Total
368 euros 62 62 3 8 = 496 22,816 euros 3.58 euros
349 euros 33 33 3 10 = 330 11,517 euros 3.58 euros
826 34,333 euros 2,950 euros
Unit purchase cost No. purchased per annum No. uses per annum Cost per annum Per instrument sterilization cost* Total
37,283 euros
Sterilization of endoscopic accessories/total sterilization workload of the sterilization unit: 0.37%. Annual personnel costs: 566,207 euros for 2,942 workdays, 8 h on duty/d, including 6 h of actual sterilization work of 17,652 h or 1,059 120 min. Direct annual personnel costs: 424,655 euros Cost of the min. at work: 0.40 euros (all classes of personnel mixed). Worktime per instrument: 4.7 min. Indirect annual personnel costs: 141,552 euros. Miscellaneous consumables: 99,909 euros. Costs of energy: 16,029 euros. Depreciation expense (reprocessors and unit equipment): 116,210 euros. *4.7 * 0.40 + (((116 210 + 16 029 + 99 909 + 141 552) * 0.37)/824) = 3.58 euros. Annual cost for strategy 1: 62 * 368 + 33 * 349 + 824 * 3.58 = 37,283 euros. B) Strategy 2: externalized reprocessing Reusable double-lumen sphincterotome
Retrieval basket
Total
368 euros 62 62 3 8 = 496 22,816 euros 7 euros
349 euros 33 33 3 10 = 330 11,517 euros 7 euros
826 34,333 euros 5,768 euros
Unit purchase cost No. purchased per annum No. uses per annum Cost per annum Per instrument sterilization cost* Total
40,101 euros
Unit reprocessing cost by third party: 7 euros. Annual cost for strategy 2: 62 * 368 + 33 * 349 + 824 * 7 = 40,101 euros. C) Strategy 3: single-use instruments only Reusable double-lumen sphincterotome
Retrieval basket
Total
125 euros 494 494
162 euros 330 330
824
61,750 euros
53,460 euros
115,210 euros
Unit purchase cost No. purchased per annum No. uses per annum Cost per annum
Annual cost of strategy 3: 494 * 125 = 330 * 162 = 115 210 euros.
pathogens, both were considered incidental contaminants. Tests for HBV and HCV were negative for 100% of devices tested. Number of uses per device The median and mean numbers of uses per type of device are presented in Table 2. The median number of safe and technically satisfactory reuses was similar for all types of devices tested (between 8 and 12 uses per device; Table 2). The most common reasons for VOLUME 60, NO. 2, 2004
discarding retrieval baskets were worn or twisted wires, which made it difficult to grasp stones, and mechanical lithotripsy, which led to destruction of 9 of 20 baskets. Sphincterotomes frequently were discarded because of a break in the cutting wire (13/30) and folding or twisting of the catheter or wire (12/30); more specific for the double-lumen sphincterotome was occlusion of the injection/wire port. Until the devices were discarded, there was no evidence of a decrease in technical quality through consecutive uses. GASTROINTESTINAL ENDOSCOPY
249
F Prat, J-F Spieler, S Paci, et al.
ERCP ancillary devices: reliability, cost-effectiveness, and safety of reuse
Table 2. Number of efficient uses per type of reusable ERCP device
N
Median No. uses
Range
Sphincterotome single-lumen type (KD-5q)
10
12
1-44
Sphincterotome single-lumen type 2 (KD-21q)
10
10
2-22
Double-lumen sphincterotome (KD-6G-12q1)
10
8
2-14
Retrieval basket
20
10
1-36
Specific reason for dysfunction or destruction
Mean No. uses (SD) [95% CI] 16.5 (14.8) [7.8, 25.7] 11.7 (8.4) [6.0, 16.9] 7.7 (3.8) [5.3, 10.1] 13.2 (9.2) [9.2, 17.2]
None 6/10, cutting wire break 6/10, Injection port occlusion 9/20, destruction by mechanical lithotripsy
SD, Standard deviation; CI, confidence interval.
Table 3. Detailed costs of in-house sterilization Elements of cost
Data
No. sterilized accessories
Work time per accessory
Amount in euros
894 —62 sphincterotomes used 8 times each —33 retrieval baskets used 10 times each 4.7 minutes
Manual direct personnel costs Manual indirect personnel costs Miscellaneous (consumables, maintenance, etc.) Utilities Depreciation expense
424,655 141,552 99,909
16,029 116,210
Per instrument sterilization cost
3.58
Figure 1. Cost of different strategies for ERCP devices according to number of procedures performed annually. 1, in-house reprocessing; 2, externalized reprocessing; 3, single-use only.
Cost analysis The overall volume of devices sterilized in 2002 in the hospital was 1460 m3 of which 0.37% were endoscopic accessories. There were 494 ERCPs performed in the endoscopy unit in 2002. Strategy 1 was the least costly, at V37,283 per year compared with strategy 2 (V40,101 per year) and with strategy 3 (V115,210 per year), which was the most expensive, regardless of the number of procedures performed. Details of the cost analysis for the 3 strategies are presented in Tables 1 and 3. The variation in cost for each strategy as a function of the number of procedures performed annually is depicted in Figure 1. DISCUSSION Previous studies of the reuse of endoscopic devices have focused on the reuse of biopsy forceps,9-12 cost savings resulting from using reusable5,6 or repro250
GASTROINTESTINAL ENDOSCOPY
cessed single-use devices,7,13 and in vitro microbial contamination after reprocessings.14 The present clinical study assessed the safety and the costs of performing ERCP with reusable accessory devices in a sufficiently large number of cases. Reprocessing of single-use devices was not evaluated, although several studies have demonstrated this approach to be feasible and cost saving.2,7 Single-use devices generally cannot be steam sterilized and must be gas sterilized (ethylene oxide), which is considered less reliable, especially for devices with long and thin inner channels. Moreover, although a manufacturer may label a device as single use for marketing purposes instead of strict medical reasons, use of the device contrary to the manufacturer’s recommendations is discouraged in countries where this would potentially increase the legal liability of the physician and/or institution.15 VOLUME 60, NO. 2, 2004
ERCP ancillary devices: reliability, cost-effectiveness, and safety of reuse
Routine surveillance cultures of endoscopic accessories, which significantly increases costs, is not common. However, as demonstrated by the present study, strict adherence to current reprocessing guidelines makes reuse of therapeutic ERCP devices reasonably safe. Because the risk for transmission of HCV infection through digestive endoscopy has been of concern in recent years (especially when biopsy specimens are obtained or therapeutic procedures performed), the present study was carefully designed to detect hepatitis virus contamination of devices, in addition to a broad spectrum of infectious agents.16 Only 11 patients (based on the presence of serum antibodies), were found to be potentially capable of HCV transmission. The presence of species reported to have occasionally caused cross-contamination during endoscopy, such as H pylori and Campylobacter species, also was evaluated.17 Variant Creutzfeldt-Jakob disease, which is caused by prions, is the only type of infection that cannot be definitely precluded. However, high-temperature steam sterilization is considered effective against prions, and the actual risk of endoscopy-related transmission is considered extremely low, if not negligible, for this species, except perhaps for biopsy specimens taken from the terminal ileum.18 Canard et al.19 found in a comparative study that single-lumen reusable sphincterotomes were as effective as triple-lumen single-use instruments in terms of success rates and procedure-related complications. In the present study, both sphincterotomes and retrieval baskets functioned effectively after more than 30 reuses. However, the range of reuses per instrument was variable; wire breaks were observed after the first sphincterotomy with a few sphincterotomes, and baskets were destroyed when mechanical lithotripsy was performed. Single-use instruments allegedly eliminate the risk of infection and are presumably technically optimal at the time of use, but disposal of these devices increases the environmental impact of medical waste, and their cost-effectiveness is questionable. The results of comparative cost analyses for reusable vs. single-use biopsy forceps in various settings are variable.9-12 However, the purchase price of a single-use forceps is now only slightly higher than the estimated per-use plus reprocessing cost of reusable forceps. Thus, single-use forceps are now generally recommended for routine upper and lower endoscopy. The costs for reprocessing ERCP devices are the same as for biopsy forceps. Kozarek et al.7 estimated the cost of in-house ethylene-oxide reprocessing of single-use sphincterotomes at $6.00. Most endoscopic centers are components of a larger multidisciplinary institution that has a sterilization VOLUME 60, NO. 2, 2004
F Prat, J-F Spieler, S Paci, et al.
unit that is used mostly to support surgical services. The sterilization of endoscopy devices, therefore, usually represents only a minor fraction of the total workload. However, actual data on the cost of sterilization of endoscopy accessories are scarce, and, therefore, a method was developed, based on a published health-care analysis,20 to analyze the contribution of interventional endoscopy to the overall workload of a sterilization unit. Although ERCP represented only 0.37% of the overall sterilization activity in our hospital, the volume of devices sterilized was sufficient to make it efficient and cost effective. A recent internal report of the Assistance Publique-Hoˆpitaux de Paris estimated that 1,000,000 items are sterilized per year per institution in France.21 Although results will vary, depending on local social and economic conditions, the method of cost analysis developed for the present study can be used in any institution to determine costs. The cost of external reprocessing (strategy 2) was comparable with that of in-house reprocessing, but the cost of the former may vary, depending on the level of competition among vendors in the local market; in a highly competitive market, the cost may be less than that for in-house reprocessing. Two factors not considered in the present analysis could increase the economic disadvantage of a single-use policy: the cost of medical waste disposal obviously is higher for disposable compared with reusable items; and costs could be reduced by reusing single-lumen sphincterotomes, which are cheaper and more durable than doublelumen sphincterotomes. A substantial decrease in the cost of single-use devices could significantly narrow the cost advantage of reusable devices. Moreover, disposable items are preferred when they offer a technical advantage or when reprocessing is impossible or unsatisfactory (e.g., guidewires, dilators, balloons). Institutions that do not have largescale sterilization activities and/or an extremely low volume of ERCPs, may find disposable accessories to be preferable, regardless of relative cost issues.22 However, the results of the present study show that reprocessing of reusable ERCP accessories is reliable and cost effective, whether reprocessing is done within or external to the institution, when compared with the use of single-use devices. DISCLOSURE This study was financially supported by Olympus Optical Co., Ltd, Tokyo, Japan. REFERENCES 1. Provenzale D. Economic analysis of endoscopic accessories. Gastrointest Endosc Clin N Am 1999;9:573-86. GASTROINTESTINAL ENDOSCOPY
251
F Prat, J-F Spieler, S Paci, et al.
ERCP ancillary devices: reliability, cost-effectiveness, and safety of reuse
2. Petersen BT. Advantages of disposable endoscopic accessories. Gastrointest Endosc Clin N Am 2000;10:341-8. 3. Wolfsen HC. Advantages of reusable accessories. Gastrointest Endosc Clin N Am 2000;10:349-59. 4. Furman PJ. Third-party reprocessing of endoscopic accessories. Gastrointest Endosc Clin N Am 2000;10:385-92. 5. Lee RM, Vida F, Kozarek RA, Raltz SL, Ball TJ, Patterson DJ, et al. In vitro and in vivo evaluation of reusable doublechannel sphincterotome. Gastrointest Endosc 1999;49: 477-82. 6. Cohen J, Haber GB, Kortan P, Dorais AL, Scheider DM, Cirocco M, et al. A prospective study of the repeated use of sterilized papillotomes and retrieval baskets for ERCP: quality and cost-analysis. Gastrointest Endosc 1997;45:122-7. 7. Kozarek RA, Raltz SL, Ball TJ, Patterson DJ, Brandabur JJ. Reuse of disposable sphincterotomes for diagnostic and therapeutic ERCP: a one-year prospective study. Gastrointest Endosc 1999;49:39-42. 8. Axon A, Kruse A, Urgell R, Struelens M, Petersen C, Spencer K, et al. ESGE guidelines for the reprocessing of accessories in digestive endoscopy. Endoscopy 1996;28:534-5. 9. Yang R, Ng S, Nichol M, Laine L. A cost and performance evaluation of disposable and reusable biopsy forceps in GI endoscopy. Gastrointest Endosc 2000;51:266-70. 10. Deprez PH, Horsmans Y, Van Hassel M, Hoang P, Piessevaux H, Geubel A. Disposable versus reusable biopsy forceps; a prospective cost evaluation. Gastrointest Endosc 2000;51: 262-5. 11. Rizzo J, Bernstein D, Gress F. A performance, safety and cost comparison of reusable and disposable endoscopic biopsy forceps: a prospective, randomized trial. Gastrointest Endosc 2000;51:257-61. 12. Lejeune C, Prost P, Michiels C, Roullaud-Guenfoudi MP, Phelip JM, Martin L, et al. Disposable versus reusable biopsy forceps. A prospective cost-analysis in the gastrointestinal
252
GASTROINTESTINAL ENDOSCOPY
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
endoscopy unit of the Dijon University Hospital [French]. Gastroenterol Clin Biol 2001;25:669-73. Carr-Locke DL, Conn MI, Faigel DO, Laing K, Leung JW, Mills MR, et al. Device reprocessing companies. ASGE technology status evaluation, May 1998. Jung M, Beilenhoff U, Pietsch M, Kraft B, Rippin G. Standardized reprocessing of reusable colonoscopy biopsy forceps is effective: results of a German multicenter study. Endoscopy 2003;35:197-202. Wilkinson M, Simmons N, Bramble M, Leicester R, D’Silva J, Boys R, et al. Report of the working party of the endoscopy committee of the British Society of Gastroenterology on the reuse of endoscopic accessories. Gut 1998;42:304-6. Bronowicki JP, Venard V, Botte´ C, Monhoven N, Gastin I, Chone L, et al. Patient-to-patient transmission of hepatitis C virus during colonoscopy. N Engl J Med 1997;337:237-40. Langenberg W, Rauws EA, Oudbier JH, Tytgat GN. Patientto-patient transmission of Campylobacter pylori by fiberoptic gastroduodenoscopy and biopsy. J Infect Dis 1990;161: 507-11. Axon ATR, Beilenhoff U, Bramble MG, Ghosh S, Kruse A, Mc Donnell GE, et al. Variant Creutzfeldt-Jakob disease (vCJD) and gastrointestinal endoscopy. Endoscopy 2001;33:1070-80. Canard JM, Cellier C, Houcke P, Laurent J, Gorce D, Landi B. Prospective multicenter study comparing a standard reusable sphincterotome with a disposable triple-lumen sphincterotome. Gastrointest Endosc 2000;51:704-7. Drummond M, O’Brien B, Stoddart G, Torrance G. Methods for the economic evaluation of health care programmes. 2nd ed. Oxford: University Press; 1997. Lajonche`re JP. Report of the study group on sterilization of medical devices. Internal report. Paris: Assistance PubliqueHtptaux de Paris; August 2002. Kozarek RA. Coming clean on reuse of endoscopic equipment. Gut 1998;42:155-6.
VOLUME 60, NO. 2, 2004
Background: The choice between reusable and single-use devices for ERCP depends on various medical and economic criteria. This study evaluated the reliability and the safety (risk of crosscontamination) of reusable devices. A cost analysis of the use of reusable devices also was conducted. Methods: All patients referred for ERCP that required use of a sphincterotome or a retrieval basket were eligible for inclusion in a clinical study of 4 different devices (3 types of sphincterotome, 1 type of retrieval basket). All devices were steam sterilized. Before each use, each device was subjected to bacteriologic and virologic tests (hepatitis C virus, hepatitis B virus markers). Devices were examined before and after each procedure. The numbers of safe and efficient procedures that could be performed with each device were assessed. Three strategies were compared in a cost analysis: internal reprocessing (strategy 1), external reprocessing (strategy 2), and single-use (strategy 3). Inputs used were the results of the clinical study, hospital data for 1 year of endoscopic activity, and market prices. Results: A total of 342 patients underwent the following procedures: sphincterotomy (248 patients), stent insertion (59 patients), use of basket without sphincterotomy (14 patients), and diagnostic ERCP/unsuccessful cannulation (21 patients). At the time of ERCP, 36 patients had viral or bacterial infection. Fifty instruments were used (20 single-lumen sphincterotomes, 10 double lumen sphincterotomes, 20 retrieval baskets). Overall, the median number of efficient uses per device was 10. The median number of efficient uses by each type of device was the following: single-lumen sphincterotome, 12; double-lumen sphincterotome, 8; and, retrieval baskets, 10. All virologic and bacteriologic tests for all instruments were negative. The cost-optimization analysis found that strategy 1 is cost effective (e37,283/y) compared with strategy 2 (e40,101/y) and especially with Strategy 3 (e115,210/y). Conclusions: Reuse of the sphincterotomes and baskets evaluated in this study during ERCP is safe in terms of infectious hazards. Because they endure numerous uses, reusable instruments are cost effective, especially when compared with single-use accessories. (Gastrointest Endosc 2004;60:246-52.)
Interventional endoscopy contends with two conflicting requirements: optimal patient safety, by elimination of infectious risks and procedure-related complications, and control of costs in an era of decliningreimbursement,expensiveadvancesintechnology, and growing demands for health care by an aging population. Ancillary devices used in intervenReceived December 12, 2003. For revision January 12, 2004. Accepted February 24, 2004. Current affiliations: Service d’he´patogastroente´rologie, Inserm U537, Ste´rilisation centrale, Service de Microbiologie, CHU Biceˆtre, Le Kremlin-Biceˆtre, France. This work was presented in part at Digestive Diseases Week, May 20-23, 2001, Atlanta, Georgia (Gastrointest Endosc 2001;53:AB76). Reprint requests: Dr. Fre´de´ric PRAT, Service d’he´patogastroente´rologie, CHU Biceˆtre, 78 rue du Ge´ne´ral Leclerc, 94275 Le Kremlin-Biceˆtre, France. Copyright Ó 2004 by the American Society for Gastrointestinal Endoscopy 0016-5107/$30.00 PII: S0016-5107(04)01685-2 246
GASTROINTESTINAL ENDOSCOPY
tional endoscopy are representative of this dilemma: single-use devices supposedly are devoid of technical flaws and infectious agents, but economic pressure may dictate the choice for reusable devices.1-4 Market prices set by device manufacturers and regulations established by health authorities often are deciding factors in this choice. However, data from scientific studies may help to achieve an optimal balance between patient safety and cost containment. It was our hypothesis that single-use devices were, by definition, optimal and that only reusable devices required assessment in terms of safety, especially transmission of infectious agents and technical reliability. Therefore, a 2-part study was conducted: a prospective clinical evaluation of standard reusable ERCP accessories (infectious safety and reliability) and a comparative cost analysis of single-use and reusable devices by using the results of the clinical evaluation and accounting data from our hospital. VOLUME 60, NO. 2, 2004
ERCP ancillary devices: reliability, cost-effectiveness, and safety of reuse
PATIENTS AND METHODS Safety and reliability of ERCP devices Patients. Patients referred for interventional biliary endoscopy were eligible for inclusion in the study if use of a sphincterotome and/or a retrieval basket were anticipated. Procedures included were the following: endoscopic sphincterotomy, regardless of indication; stone extraction without sphincterotomy; stone extraction with a previous sphincterotomy; and diagnostic ERCP, requiring use of a sphincterotome for selective cannulation. Exclusion criteria were the following: diagnostic ERCP by using a standard catheter, therapeutic ERCP in which a single-use device was required, and a non-biliary or a pancreatic therapeutic procedure. The study was approved by our local ethical review board. Devices. Four distinct types of device were studied: two different single-lumen sphincterotomes, with a monofilament, 3-cm-long cutting wire and a short cannulation tip (KD-5q and KD-21q; Olympus Optical Co., Ltd. Tokyo, Japan); a double-lumen sphincterotome (KD-6G-12q1; Olympus); and, a medium-sized braided wire retrieval basket (FG-22q; Olympus). Hypotheses. The primary study end point was the number of times the ERCP accessory device could be used efficiently and safely. Based on a hypothesis that each type of sphincterotome could be used a mean of 7 6 2 times,5-7 the assessment of 10 items of each type (i.e., 30 sphincterotomes overall) appeared to be adequate (95% CI[5.76, 8.24]). The number of uses for retrieval baskets was a secondary end point. Two different devices (sphincterotome and retrieval basket) usually are used during most procedures, but sphincterotomes are used, overall, 1.5 times more often than retrieval baskets. Thus, the number of baskets to be studied was set at 20. A sphincterotome is the first device used in nearly 98% of the therapeutic ERCPs performed per year in our department (approximately 500 ERCPs/y). It was calculated that slightly more than 200 patients had to be enrolled in the study (30 sphincterotomes 3 7 uses = 210 patients). However, it was decided to continue the enrollment of patients until all the devices had been discarded.
Device reprocessing and microbiologic studies Reprocessing. Before the first use, each device was steam sterilized in an autoclave. The devices then were reprocessed according to guidelines established by the European Society of Gastrointestinal Endoscopy, which include 10 steps: immediate immersion and washing in fresh detergent solution after the endoscopic procedure, dismantling and flushing with detergent, brushing with a cleaning brush, US cleaning (20 minutes at 608C), rinsing and flushing under running clean water, drying, sterile packaging, transfer to the central sterilization unit of the hospital, steam sterilization (1348C for 18 minutes), and return from the sterilization unit to the endoscopy unit for storage.8 The sphincterotome handles were packed separately and steam sterilized in the same manner as sphincterotomes and baskets. VOLUME 60, NO. 2, 2004
F Prat, J-F Spieler, S Paci, et al.
Microbiologic studies. Each sterilized device was tested before use by flushing the injection port under sterile conditions with 2 mL of sterile saline solution. Samples were collected for bacteriologic and virologic studies. Briefly, one sample was used for culture in 4 different media (aerobic, anaerobic, Lowenstein’s medium for mycobacteria, and Sabouraud’s medium for yeasts and fungi), as well as specific media for Helicobacter pylori and Campylobacter species. Another sample was tested for hepatitis C virus (HCV) by quantitative polymerase chain reaction (PCR) amplification and hepatitis B virus (HBV) markers (HBs Ag and HBe Ag). The detection threshold for this HCV PCR test (Amplicor 2.0; Roche, Basel, Switzerland) is 100 copies. Patients also underwent serologic testing for HBV and HCV (HBs Ag, anti-HBs Ab for HBV, total anti-HCV immunoglobulins for HCV), unless their viral status was known. For patients who tested positive for serum viral markers, reverse transcriptase–PCR for HCV and serum HBV b-DNA also were obtained. With an expected frequency of iatrogenic infection of 0.1%: 95% CI[0.001, 0.005], 153 samples were sufficient to detect cross-contamination (data from epi2002 software). Thus, virologic and bacteriologic tests were discontinued after inclusion of 200 patients. However, this calculation reflected a broad variety of infections, but it did not take into account the number of patients actually at risk for each specific type, especially for viral species. Assessment of devices A double-tracking system was used to monitor all devices through successive ERCPs and reprocessings: each device was given an identification number, which was marked on the device with indelible ink. Records of each use, including date and patient identification, were kept in a registry and separately by using an individual form for each patient. Also recorded on the form were the diagnosis, the procedure performed, infectious risk factors (uncontrolled septicemia, positive blood cultures, positive viral markers, etc.), and the observations about the device. Each device used during a procedure was carefully examined and tested by the endoscopist immediately upon completion of the procedure. The device was graded as suitable or unsuitable for another procedure. To be judged suitable, all devices had to be fully operable. Sphincterotomes had to have a straight cutting wire, patent lumen(s), no leakage when flushed, effective bending at the tip, and an intact electrical connection. Dormia baskets had to have widely unfolding wires and smooth motion of the basket into the sheath when retracted. An instrument that did not fulfill all criteria, or if any part of it was broken, was considered unsuitable for further use and was immediately discarded. To standardize assessment among endoscopists, all devices used throughout the study were assessed by one of 3 senior endoscopists (F.P., J.F., A.D.C.). All endoscopists were blinded as to the number of previous reprocessings of each of the devices used throughout the study. The mean (standard deviation) and median numbers of uses were calculated per type of device as well as the range. GASTROINTESTINAL ENDOSCOPY
247
F Prat, J-F Spieler, S Paci, et al.
ERCP ancillary devices: reliability, cost-effectiveness, and safety of reuse
COST ANALYSIS The 1-year cost of 3 strategies for controlling costs of ERCP accessories was assessed: strategy 1 (that used in our hospital) consisted of in-house reprocessing of accessories; strategy 2 consisted of outsourcing of accessory device reprocessing to a third party; strategy 3 consisted in using only single-use devices. Each strategy generates different types of costs: for strategy 3, costs pertain essentially to purchase of the accessory devices, whereas, strategies 1 and 2 include purchase and reprocessing costs. The global cost of strategy 1 was estimated by using accounting data from our center (a 997-bed university-affiliated hospital in the greater Paris area). The overall costs for the other two strategies were determined from the number of procedures performed in our center, the market prices for the accessories used, and reprocessing costs. The comparison was based on the use of double-lumen sphincterotomes and retrieval baskets. Single-lumen sphincterotomes were not analyzed because nearly all disposable sphincterotomes currently available have two lumens. This study was approved by our local ethical review board. For purposes of the cost analysis, the median number of uses per reusable device as determined in the clinical study was taken as the number of uses per reusable instrument: this was 8 for reusable double-lumen sphincterotomes and 10 for reusable retrieval baskets. For the number of endoscopic procedures preformed, the actual number of ERCPs performed in our unit during 1 year (2002) was used. Purchase costs were the actual costs negotiated between the hospital and the manufacturers, for both reusable and single-use devices. Reprocessing costs Strategy 1 (in-house reprocessing). Reprocessing costs were estimated by using accounting data from the hospital. Sterilization takes place in a central sterilization unit, a division of the pharmacy department. The following costs were identified directly from hospital accounting records: personnel charges and wages, consumables (sterilization products, disposable items, linen, etc.), and purchase and maintenance of reprocessing equipment. Personnel costs are divided into two categories: direct costs (actual time spent working in the sterilization unit) and indirect costs (education, etc.). The sterilization unit staff includes 20 full-time nurse’s aides, 7 full-time nurses, one full-time manager, and a half-time pharmacist. Expenditures for utilities (water, electric power) are charged globally to the hospital; the relative proportion of these expenditures for the sterilization unit was calculated by using the fraction of the surface area of the hospital occupied by the sterilization unit. The cost of sterilization for endoscopic accessories was estimated by calculating a relative workload ratio for endoscopic accessories reprocessing vs. the total workload of the sterilization unit and then by applying this ratio to the costs of consumables, reprocessing equipment, and indirect personnel costs, and also to direct costs for personnel. Strategy 2 (external reprocessing). An estimate of the costs for externalized sterilization of endoscopic 248
GASTROINTESTINAL ENDOSCOPY
accessories was based on mean market prices (in euros, based on prices quoted in 2002). The cost of waste disposal was not taken into account. Detailed data used for calculations for each of the 3 strategies are presented in Table 1.
RESULTS Patients and procedures performed A total of 342 consecutive patients (144 men, 198 women; mean age 69.7 [17.2] years) were enrolled in the study. The indication for ERCP was the following: suspected choledocholithiasis, 235 patients (68.7%); cholestasis with suspected biliary stricture, 79 patients (23.1%), 12 of which were benign; chronic pancreatitis, 10 patients (2.9%); biliary leak, 8 patients (2.3%); and suspected sphincter of Oddi dysfunction, 10 patients (2.9%). Thirty-six (18%) of the 200 patients for whom virologic and bacteriologic testing was obtained were identified as potential sources for transmission of infection (cross-contamination). These patients had the following infections: chronic HCV infection (n = 11), chronic HBV infection (n = 4), HIV infection (n = 5), and septicemia with positive blood cultures (all gram-negative strains) at the time of ERCP (n = 16). Procedures performed were the following: endoscopic biliary sphincterotomy with stone extraction, 194 patients (56.7%); biliary sphincterotomy without stone extraction, 49 patients (14.3%); sphincterotomy with insertion of a biliary stent, 55 patients (16.1%); stent exchange, 4 patients (1.2%); stone extraction (previous biliary sphincterotomy), 14 patients (4.1%); pancreatic sphincterotomy, 5 patients (1.5%); and non-therapeutic ERCP (i.e., intended therapy but ductal opacification alone), 15 patients (4.4%). ERCP was unsuccessful in 6 patients (1.7%), but these procedures were included in the analysis because the study devices were used. There was no case of successful cannulation followed by unsuccessful sphincterotomy. Immediate complications were observed in 3 patients: two post-sphincterotomy hemorrhage, one duodenal perforation. Failures and complications occurred with devices in their first to 13th use. There was no correlation between failure and complications and dysfunction of the devices. Microbiologic analyses Analyses for bacterial and yeast contaminants were negative except for 4 devices. The organisms identified were coagulase-negative Staphylococcus (3/4) and hemolytic Streptococcus species (1/4). Because these organisms are not usually digestive-tract VOLUME 60, NO. 2, 2004
ERCP ancillary devices: reliability, cost-effectiveness, and safety of reuse
F Prat, J-F Spieler, S Paci, et al.
Table 1. Elements of cost calculation (costs for 1 y of endoscopic activity) A) Strategy 1: in-house reprocessing Reusable double-lumen sphincterotome
Retrieval basket
Total
368 euros 62 62 3 8 = 496 22,816 euros 3.58 euros
349 euros 33 33 3 10 = 330 11,517 euros 3.58 euros
826 34,333 euros 2,950 euros
Unit purchase cost No. purchased per annum No. uses per annum Cost per annum Per instrument sterilization cost* Total
37,283 euros
Sterilization of endoscopic accessories/total sterilization workload of the sterilization unit: 0.37%. Annual personnel costs: 566,207 euros for 2,942 workdays, 8 h on duty/d, including 6 h of actual sterilization work of 17,652 h or 1,059 120 min. Direct annual personnel costs: 424,655 euros Cost of the min. at work: 0.40 euros (all classes of personnel mixed). Worktime per instrument: 4.7 min. Indirect annual personnel costs: 141,552 euros. Miscellaneous consumables: 99,909 euros. Costs of energy: 16,029 euros. Depreciation expense (reprocessors and unit equipment): 116,210 euros. *4.7 * 0.40 + (((116 210 + 16 029 + 99 909 + 141 552) * 0.37)/824) = 3.58 euros. Annual cost for strategy 1: 62 * 368 + 33 * 349 + 824 * 3.58 = 37,283 euros. B) Strategy 2: externalized reprocessing Reusable double-lumen sphincterotome
Retrieval basket
Total
368 euros 62 62 3 8 = 496 22,816 euros 7 euros
349 euros 33 33 3 10 = 330 11,517 euros 7 euros
826 34,333 euros 5,768 euros
Unit purchase cost No. purchased per annum No. uses per annum Cost per annum Per instrument sterilization cost* Total
40,101 euros
Unit reprocessing cost by third party: 7 euros. Annual cost for strategy 2: 62 * 368 + 33 * 349 + 824 * 7 = 40,101 euros. C) Strategy 3: single-use instruments only Reusable double-lumen sphincterotome
Retrieval basket
Total
125 euros 494 494
162 euros 330 330
824
61,750 euros
53,460 euros
115,210 euros
Unit purchase cost No. purchased per annum No. uses per annum Cost per annum
Annual cost of strategy 3: 494 * 125 = 330 * 162 = 115 210 euros.
pathogens, both were considered incidental contaminants. Tests for HBV and HCV were negative for 100% of devices tested. Number of uses per device The median and mean numbers of uses per type of device are presented in Table 2. The median number of safe and technically satisfactory reuses was similar for all types of devices tested (between 8 and 12 uses per device; Table 2). The most common reasons for VOLUME 60, NO. 2, 2004
discarding retrieval baskets were worn or twisted wires, which made it difficult to grasp stones, and mechanical lithotripsy, which led to destruction of 9 of 20 baskets. Sphincterotomes frequently were discarded because of a break in the cutting wire (13/30) and folding or twisting of the catheter or wire (12/30); more specific for the double-lumen sphincterotome was occlusion of the injection/wire port. Until the devices were discarded, there was no evidence of a decrease in technical quality through consecutive uses. GASTROINTESTINAL ENDOSCOPY
249
F Prat, J-F Spieler, S Paci, et al.
ERCP ancillary devices: reliability, cost-effectiveness, and safety of reuse
Table 2. Number of efficient uses per type of reusable ERCP device
N
Median No. uses
Range
Sphincterotome single-lumen type (KD-5q)
10
12
1-44
Sphincterotome single-lumen type 2 (KD-21q)
10
10
2-22
Double-lumen sphincterotome (KD-6G-12q1)
10
8
2-14
Retrieval basket
20
10
1-36
Specific reason for dysfunction or destruction
Mean No. uses (SD) [95% CI] 16.5 (14.8) [7.8, 25.7] 11.7 (8.4) [6.0, 16.9] 7.7 (3.8) [5.3, 10.1] 13.2 (9.2) [9.2, 17.2]
None 6/10, cutting wire break 6/10, Injection port occlusion 9/20, destruction by mechanical lithotripsy
SD, Standard deviation; CI, confidence interval.
Table 3. Detailed costs of in-house sterilization Elements of cost
Data
No. sterilized accessories
Work time per accessory
Amount in euros
894 —62 sphincterotomes used 8 times each —33 retrieval baskets used 10 times each 4.7 minutes
Manual direct personnel costs Manual indirect personnel costs Miscellaneous (consumables, maintenance, etc.) Utilities Depreciation expense
424,655 141,552 99,909
16,029 116,210
Per instrument sterilization cost
3.58
Figure 1. Cost of different strategies for ERCP devices according to number of procedures performed annually. 1, in-house reprocessing; 2, externalized reprocessing; 3, single-use only.
Cost analysis The overall volume of devices sterilized in 2002 in the hospital was 1460 m3 of which 0.37% were endoscopic accessories. There were 494 ERCPs performed in the endoscopy unit in 2002. Strategy 1 was the least costly, at V37,283 per year compared with strategy 2 (V40,101 per year) and with strategy 3 (V115,210 per year), which was the most expensive, regardless of the number of procedures performed. Details of the cost analysis for the 3 strategies are presented in Tables 1 and 3. The variation in cost for each strategy as a function of the number of procedures performed annually is depicted in Figure 1. DISCUSSION Previous studies of the reuse of endoscopic devices have focused on the reuse of biopsy forceps,9-12 cost savings resulting from using reusable5,6 or repro250
GASTROINTESTINAL ENDOSCOPY
cessed single-use devices,7,13 and in vitro microbial contamination after reprocessings.14 The present clinical study assessed the safety and the costs of performing ERCP with reusable accessory devices in a sufficiently large number of cases. Reprocessing of single-use devices was not evaluated, although several studies have demonstrated this approach to be feasible and cost saving.2,7 Single-use devices generally cannot be steam sterilized and must be gas sterilized (ethylene oxide), which is considered less reliable, especially for devices with long and thin inner channels. Moreover, although a manufacturer may label a device as single use for marketing purposes instead of strict medical reasons, use of the device contrary to the manufacturer’s recommendations is discouraged in countries where this would potentially increase the legal liability of the physician and/or institution.15 VOLUME 60, NO. 2, 2004
ERCP ancillary devices: reliability, cost-effectiveness, and safety of reuse
Routine surveillance cultures of endoscopic accessories, which significantly increases costs, is not common. However, as demonstrated by the present study, strict adherence to current reprocessing guidelines makes reuse of therapeutic ERCP devices reasonably safe. Because the risk for transmission of HCV infection through digestive endoscopy has been of concern in recent years (especially when biopsy specimens are obtained or therapeutic procedures performed), the present study was carefully designed to detect hepatitis virus contamination of devices, in addition to a broad spectrum of infectious agents.16 Only 11 patients (based on the presence of serum antibodies), were found to be potentially capable of HCV transmission. The presence of species reported to have occasionally caused cross-contamination during endoscopy, such as H pylori and Campylobacter species, also was evaluated.17 Variant Creutzfeldt-Jakob disease, which is caused by prions, is the only type of infection that cannot be definitely precluded. However, high-temperature steam sterilization is considered effective against prions, and the actual risk of endoscopy-related transmission is considered extremely low, if not negligible, for this species, except perhaps for biopsy specimens taken from the terminal ileum.18 Canard et al.19 found in a comparative study that single-lumen reusable sphincterotomes were as effective as triple-lumen single-use instruments in terms of success rates and procedure-related complications. In the present study, both sphincterotomes and retrieval baskets functioned effectively after more than 30 reuses. However, the range of reuses per instrument was variable; wire breaks were observed after the first sphincterotomy with a few sphincterotomes, and baskets were destroyed when mechanical lithotripsy was performed. Single-use instruments allegedly eliminate the risk of infection and are presumably technically optimal at the time of use, but disposal of these devices increases the environmental impact of medical waste, and their cost-effectiveness is questionable. The results of comparative cost analyses for reusable vs. single-use biopsy forceps in various settings are variable.9-12 However, the purchase price of a single-use forceps is now only slightly higher than the estimated per-use plus reprocessing cost of reusable forceps. Thus, single-use forceps are now generally recommended for routine upper and lower endoscopy. The costs for reprocessing ERCP devices are the same as for biopsy forceps. Kozarek et al.7 estimated the cost of in-house ethylene-oxide reprocessing of single-use sphincterotomes at $6.00. Most endoscopic centers are components of a larger multidisciplinary institution that has a sterilization VOLUME 60, NO. 2, 2004
F Prat, J-F Spieler, S Paci, et al.
unit that is used mostly to support surgical services. The sterilization of endoscopy devices, therefore, usually represents only a minor fraction of the total workload. However, actual data on the cost of sterilization of endoscopy accessories are scarce, and, therefore, a method was developed, based on a published health-care analysis,20 to analyze the contribution of interventional endoscopy to the overall workload of a sterilization unit. Although ERCP represented only 0.37% of the overall sterilization activity in our hospital, the volume of devices sterilized was sufficient to make it efficient and cost effective. A recent internal report of the Assistance Publique-Hoˆpitaux de Paris estimated that 1,000,000 items are sterilized per year per institution in France.21 Although results will vary, depending on local social and economic conditions, the method of cost analysis developed for the present study can be used in any institution to determine costs. The cost of external reprocessing (strategy 2) was comparable with that of in-house reprocessing, but the cost of the former may vary, depending on the level of competition among vendors in the local market; in a highly competitive market, the cost may be less than that for in-house reprocessing. Two factors not considered in the present analysis could increase the economic disadvantage of a single-use policy: the cost of medical waste disposal obviously is higher for disposable compared with reusable items; and costs could be reduced by reusing single-lumen sphincterotomes, which are cheaper and more durable than doublelumen sphincterotomes. A substantial decrease in the cost of single-use devices could significantly narrow the cost advantage of reusable devices. Moreover, disposable items are preferred when they offer a technical advantage or when reprocessing is impossible or unsatisfactory (e.g., guidewires, dilators, balloons). Institutions that do not have largescale sterilization activities and/or an extremely low volume of ERCPs, may find disposable accessories to be preferable, regardless of relative cost issues.22 However, the results of the present study show that reprocessing of reusable ERCP accessories is reliable and cost effective, whether reprocessing is done within or external to the institution, when compared with the use of single-use devices. DISCLOSURE This study was financially supported by Olympus Optical Co., Ltd, Tokyo, Japan. REFERENCES 1. Provenzale D. Economic analysis of endoscopic accessories. Gastrointest Endosc Clin N Am 1999;9:573-86. GASTROINTESTINAL ENDOSCOPY
251
F Prat, J-F Spieler, S Paci, et al.
ERCP ancillary devices: reliability, cost-effectiveness, and safety of reuse
2. Petersen BT. Advantages of disposable endoscopic accessories. Gastrointest Endosc Clin N Am 2000;10:341-8. 3. Wolfsen HC. Advantages of reusable accessories. Gastrointest Endosc Clin N Am 2000;10:349-59. 4. Furman PJ. Third-party reprocessing of endoscopic accessories. Gastrointest Endosc Clin N Am 2000;10:385-92. 5. Lee RM, Vida F, Kozarek RA, Raltz SL, Ball TJ, Patterson DJ, et al. In vitro and in vivo evaluation of reusable doublechannel sphincterotome. Gastrointest Endosc 1999;49: 477-82. 6. Cohen J, Haber GB, Kortan P, Dorais AL, Scheider DM, Cirocco M, et al. A prospective study of the repeated use of sterilized papillotomes and retrieval baskets for ERCP: quality and cost-analysis. Gastrointest Endosc 1997;45:122-7. 7. Kozarek RA, Raltz SL, Ball TJ, Patterson DJ, Brandabur JJ. Reuse of disposable sphincterotomes for diagnostic and therapeutic ERCP: a one-year prospective study. Gastrointest Endosc 1999;49:39-42. 8. Axon A, Kruse A, Urgell R, Struelens M, Petersen C, Spencer K, et al. ESGE guidelines for the reprocessing of accessories in digestive endoscopy. Endoscopy 1996;28:534-5. 9. Yang R, Ng S, Nichol M, Laine L. A cost and performance evaluation of disposable and reusable biopsy forceps in GI endoscopy. Gastrointest Endosc 2000;51:266-70. 10. Deprez PH, Horsmans Y, Van Hassel M, Hoang P, Piessevaux H, Geubel A. Disposable versus reusable biopsy forceps; a prospective cost evaluation. Gastrointest Endosc 2000;51: 262-5. 11. Rizzo J, Bernstein D, Gress F. A performance, safety and cost comparison of reusable and disposable endoscopic biopsy forceps: a prospective, randomized trial. Gastrointest Endosc 2000;51:257-61. 12. Lejeune C, Prost P, Michiels C, Roullaud-Guenfoudi MP, Phelip JM, Martin L, et al. Disposable versus reusable biopsy forceps. A prospective cost-analysis in the gastrointestinal
252
GASTROINTESTINAL ENDOSCOPY
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
endoscopy unit of the Dijon University Hospital [French]. Gastroenterol Clin Biol 2001;25:669-73. Carr-Locke DL, Conn MI, Faigel DO, Laing K, Leung JW, Mills MR, et al. Device reprocessing companies. ASGE technology status evaluation, May 1998. Jung M, Beilenhoff U, Pietsch M, Kraft B, Rippin G. Standardized reprocessing of reusable colonoscopy biopsy forceps is effective: results of a German multicenter study. Endoscopy 2003;35:197-202. Wilkinson M, Simmons N, Bramble M, Leicester R, D’Silva J, Boys R, et al. Report of the working party of the endoscopy committee of the British Society of Gastroenterology on the reuse of endoscopic accessories. Gut 1998;42:304-6. Bronowicki JP, Venard V, Botte´ C, Monhoven N, Gastin I, Chone L, et al. Patient-to-patient transmission of hepatitis C virus during colonoscopy. N Engl J Med 1997;337:237-40. Langenberg W, Rauws EA, Oudbier JH, Tytgat GN. Patientto-patient transmission of Campylobacter pylori by fiberoptic gastroduodenoscopy and biopsy. J Infect Dis 1990;161: 507-11. Axon ATR, Beilenhoff U, Bramble MG, Ghosh S, Kruse A, Mc Donnell GE, et al. Variant Creutzfeldt-Jakob disease (vCJD) and gastrointestinal endoscopy. Endoscopy 2001;33:1070-80. Canard JM, Cellier C, Houcke P, Laurent J, Gorce D, Landi B. Prospective multicenter study comparing a standard reusable sphincterotome with a disposable triple-lumen sphincterotome. Gastrointest Endosc 2000;51:704-7. Drummond M, O’Brien B, Stoddart G, Torrance G. Methods for the economic evaluation of health care programmes. 2nd ed. Oxford: University Press; 1997. Lajonche`re JP. Report of the study group on sterilization of medical devices. Internal report. Paris: Assistance PubliqueHtptaux de Paris; August 2002. Kozarek RA. Coming clean on reuse of endoscopic equipment. Gut 1998;42:155-6.
VOLUME 60, NO. 2, 2004