Almost every gynecologist is aware of the approaches to effective and safe abdominal, vaginal and laparoscopic hysterectomy and should also be aware of the correct indications for performing each of these procedures. However, there is a great difference in the proportions of these hysterectomy types worldwide (Garry 1998). Approximately 70-80% of hysterectomies have been performed abdominally in the United States (Easterday et al. 1983), the United Kingdom (Garry 1998), and in Finland (Vuorma et al. 1998), but only 30% in Austria (Gitsch et al. 1991). In one center in France 80% of hysterectomies have been performed vaginally (Cosson et al. 1997), while in another center, 80% of hysterectomies in nulliparous women have been performed laparoscopically (Chapron et al. 1996). These wide variations between different countries, between different units and indeed between different gynecologists, indicate that after more than 100 years' experience of hysterectomy, there is no worldwide consensus how to perform a hysterectomy in different situations. The choice of method depends more upon the experience and biases of the gynecologist than upon a critical evaluation of the operative and outcome data (Dorsey et al. 1995, Garry 1998).

Every new surgical technique leads to a learning curve. A clear shortening of operating time and to some extent also, decreasing complications with experience, have been noticed in laparoscopic hysterectomies (Ikhena et al. 1999). Laparoscopic hysterectomy was safe in experienced hands with a low complication rate in our personal series of first laparoscopic hysterectomies, even in the learning period. The operating time was reduced to a half after 80 procedures and a plateau was achieved after 60 operations, compared with the Belcohyst study, where a plateau was seen after 40 procedures (Deprest et al. 1996). Both studies were performed soon after introduction of this new and unfamiliar technique and at the beginning several operative techniques were tested before surgery was standardized. Another study showed a plateau as early as after 10 TLH with an experienced surgeon who was already familiar with laparoscopic hysterectomy technique (Rosen et al. 1998). On the other hand, in another study, no shortening of operating time was seen after the first 30 operations during the learning period in one center (Angle et al. 1995). In the two-years survey of 1165 laparoscopic hysterectomies, which covered all hospitals in Finland and not only specialized centers, it was seen that this new procedure was quickly adopted by several hospitals. The average operating time, hospital stay, convalescence time, total complication rate (10.2%) and overall major complication rate (3.5%) were similar but the rate of ureteral injuries was higher (1.3%) than reported in review articles mainly concerning experts in laparoscopy (0.3%) (Garry and Phillips 1995, Munro and Deprest 1995, Harris and Daniell 1996, Meikle et al. 1997). On the other hand, even higher ureteral injury rates (2% and 4.3%) have been reported when gynecologists have been at the beginning of their learning period (Garry 1998, Tamussino et al. 1998). Overall, the nationwide incidence of ureteral injuries decreased from 1.9% in 1993 to 0.7% in 1998 and major complications of laparoscopic hysterectomy also decreased in Finland. No long-term outcome of laparoscopic hysterectomy was analyzed, but the forthcoming years will certainly yield information about possible difference between the alternative types of hysterectomy.

Laparoscopic hysterectomy offers many advantages over abdominal hysterectomy in published randomized studies (Nezhat et al. 1992, Phipps and Nayak 1993, Raju and Auld 1994, Langebrekke et al. 1996, Olsson et al. 1996, Summitt et al. 1998, Yuen et al. 1998, Falcone et al. 1999, Marana et al. 1999).On the other hand, no real advantage has been found in randomized studies when comparing laparoscopic with vaginal hysterectomy (Summitt et al. 1992, Richardson et al. 1995), but laparoscopy offers a better view when performing hysterectomy with salpingo-oophorectomy, or in the presence of endometriosis or adhesions (Wood and Mahet 1997). In our study a somewhat longer operating time but less operative blood loss, and a shorter hospital stay and convalescence time were discovered after laparoscopic compared with abdominal hysterectomy. In addition, a clear elevation in the concentrations of biochemical tissue trauma markers was noticed after both types of hysterectomy but the increase was more marked after abdominal hysterectomy, suggesting more tissue trauma. This finding confirms the results of one earlier report (Yuen et al. 1998). However, in another study no tissue trauma differences between these two types of hysterectomy were found (Ellström et al. 1996). In that study the operating time was one hour longer in the laparoscopy group and it has been shown that the length of an operation is associated with increases in the concentrations of acute phase proteins (Ohzato et al. 1992).

Great variations in the length of hospital stay and convalescence time were seen in Finland, as also noticed between different countries. These parameters have been longest in abdominal hysterectomy but almost equal in laparoscopic and vaginal hysterectomy (Summitt et al. 1992, Johns et al. 1995, Munro and Deprest 1995, Deprest et al.1996, Meikle et al. 1997). A size of the scar, postoperative pain and tissue trauma may be the most important factors affecting the convalescence time. The mean hospital stay was 1.3 days compared with 3.3 days and sick leave was 10.9 days compared with 17.9 days in a personal series and in a national survey of laparoscopic hysterectomy, respectively. In our randomized study the convalescence time was 21.4 days, when the women themselves estimated the recovery time after which they were able to return to normal work. These figures are usually defined according to surgical tradition and are not based on critical evaluation of surgical outcome, and therefore need re-evaluation.

Two reliable Finnish registers covering the whole country were used to study the nationwide incidences of laparoscopic complications. Other studies of complications have been based on reports concerning specialized centers (Chapron et al. 1998) or surveys with low response rates (Hulka et al. 1995, Peterson et al. 1993, Levy et al. 1994), and another nationwide survey from the Netherlands included 55% of hospitals (Jansen et al. 1997). In Finland, major complications of laparoscopy were low in diagnostic and sterilization laparoscopies but they increased steeply from 1990 to 1996 in operative laparoscopies. Eighty-two percent of major complications occurred in operative laparoscopies and 69% of these in laparoscopic hysterectomies. The growing tendency towards major complications in advanced laparoscopy has also been noticed by others (Hulka et al. 1995, Jansen et al. 1997, Chapron et al. 1998), as more difficult procedures are carried out laparoscopically. The mortality rate (1/100,000) in laparoscopic procedures was the same as mortality after gynecological surgery in Finland (Virtanen et al. 1995b).

As laparoscopic hysterectomies were the main source of laparoscopic ureteral complications in Finland, urinary tract injuries associated with all types of hysterectomy were compared. The rate of ureteral injury was low in connection with vaginal, supracervical abdominal and total abdominal hysterectomy, but high in laparoscopic hysterectomy. The incidence associated with abdominal hysterectomy (0.04%) was 5 times lower and the incidence in vaginal hysterectomy (0.02%) was 2.5-5 times lower than reported in the literature (Dicker et al. 1982, Harris and Daniell 1996). In contrast, the incidence of ureteral injury during laparoscopic hysterectomy (1.4%) was 4.5 times higher than reported in review articles concerning experts (0.3%) (Garry and Phillips 1995, Munro and Deprest 1995, Harris and Daniell 1996, Meikle et al. 1997). However, the rate of ureteral injuries in association with laparoscopic hysterectomy in the whole country appears to be high but it is decreasing. Bladder injuries were also rare during vaginal (0.02%) and supracervical hysterectomy (0.03%), and more common during total abdominal hysterectomy (0.1%), but most common during laparoscopic hysterectomy (0.9%). Again, the incidence rates in association with traditional hysterectomies were lower but those of laparoscopic hysterectomies were the same as in review articles (Garry and Phillips 1995, Munro and Deprest 1995, Harris and Daniell 1996, Meikle et al. 1997). The rate of bladder injuries decreased during the study years but some minor bladder injuries, repaired laparoscopically during the primary operation, may not have been reported as they are not always compensated. According to the national register of laparoscopic hysterectomies in Finland in 1993-1994 and the Finhyst study in 1996 (Johansson, unpublished data), all major complications except some simple bladder perforations occurring in connection with laparoscopic hysterectomies were reported to the Patient Insurance Association. Complications occurring in connection with traditional hysterectomies may be under-reported because a higher number of urinary tract injuries were described in the Finhyst study in 1996 (Johansson, unpublished data) than were reported the Patient Insurance Association in the same year. Also in another Finnish study, 0.17% urinary tract injury rate and a 0.1% ureteral injury rate have been reported after total abdominal hysterectomy for benign reasons in 1983-1992 (Virtanen et al.1995a). However, the accuracy of the data from the Patient Insurance Association has not been validated and some complications may not have been reported because of improper information of the patient.

We have demostrated that the laparoscopic method of hysterectomy can be carried out at a national level and can result in faster and less painful recovery but may result in serious complications. Benefits as regards the size of scars, postoperative pain and recovery times will be irrelevant if the ureteral complication rates remain high (Garry 1998). Laparoscopic subtotal hysterectomy (Donnez et al. 1997) has been presented as being associated with fever complications, but the ureters are still very close to the point at which the uterus is separated from the cervix. Visualization or dissection of the ureters (Reich et al. 1993) during the operation is essential. In most cases, electrocoagulation was the main cause of ureteral injury in Finland, as a result of a thermal effect. Ultrasonic coagulation with an ultrasonically activated scalpel may cause less lateral thermal damage (McCarus 1996) than electrocoagulation (Phipps 1994). A laparoscopic suturing technique (Reich et al. 1993) may reduce the risk of ureteral injury but it requires practice and is time-consuming. Stapling devices offer speed but are expensive and are also associated with ureteral injuries (Woodland 1992). The use of ureteral catheters has been recommended to prevent ureteral injuries (Phipps 1995) but they have been also associated with significant morbidity (Kadar 1995). Hence, one way to decrease ureteral injuries would be to cut the uterine vessels vaginally (Mencaglia et al. 1994), especially during the learning phase. In Finland, only three of 111 ureteral injuries did occur when uterine vessels were ligated vaginally.

If ureteral or bladder injury is suspected during surgery, ureteral stenting or intravenous indigo carmine injection can be carried out. Recently, laparoscopic ultrasonographic examination has been studied during gynecologic surgery and it may offer another way to diagnose ureteral injuries, even intraoperatively (Helin et al.1998, Helin-Martikainen et al. 1998). If ureteral injury is suspected postoperatively the best diagnostic method is excretory urography and it should be carried out without hesitation.

When a new surgical technique is introduced, it is always a challenge to learn it without causing harm to patients. Patients´ expectations are also high when coming to minimal access surgery and they easily seak compensations if a complication occurs. Whereas, a similar injury occurring during open surgery is accepted as inevitable. When risks of complications in gynecological laparoscopy have been predicted it has appeared that the difficulty of the procedure is the strongest predictor of complications (Mirhashemi et al. 1998). Further, the rate of complications can be decreased by additional post-residency training courses and teamwork and an inverse correlation can be seen between a surgeon's complication rate and the number of laparoscopies performed (See et al.1993). In Finland, ureteral injuries tended to be more common in local hospitals, where the expertise is not as extensive as in university hospitals. Residents in training programs usually become skilled at advanced laparoscopic surgery, but personnel at smaller hospitals must acquire their skills by themselves. Hence, a very important aspect is proper training by means of videotapes, courses, visiting experts and at the beginning, operations supervised by experienced colleagues, before starting to perform procedures independently (Gates 1997, Chapron et al. 1997b, Garry 1998). In Finland, this means that university hospitals should arrange post-residency training, allow colleagues from smaller hospitals to participate in operations and take the responsibility to ensure that the new surgical technique is safely adopted. Otherwise, we will follow the Unites States, where in response to a high rate of complications resulting from the rapid adoption of laparoscopic cholecystectomy, The New York State Health Department made it mandatory that a surgeon should show adequate skill in at least 15 supervised cases before being allowed to perform the operations independently (Gates 1997).

Since the lifetime risk of hysterectomy is 30% in the United States, 20% in the United Kingdom (Vessey et al. 1992) and Finland (Vuorma et al. 1998) and 10% in Denmark (Settnes and Jorgensen 1996), it is very important to choose a safe, effective, but economical method of hysterectomy. Hysterectomy rates have already decreased in the United States (Weber and Lee 1996) and they have been stable in Finland in the 1990s. The incidence has decreased among young women but has increased among postmenopausal women, as a result of fibroids and uterine bleeding after increasing use of estrogen replacement therapy (Vuorma et al. 1998). Laparoscopic hysterectomy has had an influence on hysterectomy patterns in the last decade. The proportion of abdominal hysterectomies has decreased and those of laparoscopic and vaginal hysterectomies have increased in the United States (Harris and Olive 1994, Johns et al. 1995, Weber and Lee 1996), Australia (Wood et al. 1997) and Finland, but the changes have been minor in the United Kingdom (Hill et al. 1998). An important trend in the next millennium could be to avoid abdominal hysterectomy and change the pattern to laparoscopic and vaginal hysterectomies. As these approaches are less painful, are associated with a more rapid recovery and are even cost-effective when using reusable instruments, they appear to be the preferred methods (Garry 1998). The patient's symptoms as well as the gynecologist's skills influence the choice between these two methods, but everybody should have experience of all techniques of hysterectomy to offer the best treatment to the patient.