There has been a significant accumulation of collective experience with percutaneous biliary drainage (PCD) during the past 20 years. As experience with the technique has increased, its role has undergone a series of redefinitions, although early enthusiasm for percutaneous drainage has been tempered by the realities of numbers and statistics.
The first major debate arose regarding the usefulness of PCD prior to surgical decompression. In 1978, Nakayama et al , in an uncontrolled study, reported a reduction in operative and postoperative mortality from 28.3% to 8.2% when preoperative PCD was performed. Then in 1981, Denning et al  showed that operative and postoperative complications were reduced if patients underwent percutaneous drainage prior to surgery.
In that same year, based on an analysis of 155 patients who underwent biliary tract surgery, Pitt et al  established 5 clinical and 10 laboratory parameters as risk factors that could affect outcome. The most significant of these were age above 60 years, malignancy, signs of infection, jaundice, elevation of creatinine, and reduction of albumin and hematocrit. In patients with more than seven risk factors, surgical mortality was 100%. However, obstructive jaundice was only one of the risk factors. And in 1984, Gundry et al  found a reduction in mortality from 20% to 4%, and in major morbidity from 52% to 8%, when percutaneous drainage was used.
In that same year, on the other hand, in a multicenter study, Passariello et al (personal communication, 1984) showed that PCD does not confer any advantage with regard to operative and postoperative complications.
It is important to note, however, that the number of patients in all the published series is small, and the severity of the illness at the time of referral varies greatly. Therefore, comparison of risks vs benefits is uncertain at best.
Significant advances in the relatively less invasive endoscopic retrograde cholangiopancreatography (ERCP) techniques and the advent of ERCP-placed stents have decreased the number of patients referred for percutaneous drainage. The technique of ERCP stenting, initially described by Soehendra et al, has become a major therapeutic modality for the palliative decompression of malignant obstruction of the bile ducts. Stents up to 12 French in diameter, and even metallic endoprostheses, can be placed with this technique. Currently, the major indications for ERCP stenting are malignancies obstructing the extrahepatic bile ducts below the level of the bifurcation, either primarily or secondarily.
Shortcomings of both percutaneous and endoscopic stent placement include stent occlusion or migration, sepsis, abscess, hemobilia, and pseudoaneurysm formation. The incidence of these complications is comparable for both techniques.
The development of large-bore, metallic, expandable endoprostheses has brought new insights into the role of percutaneous biliary drainage. The major advantages of the metal stents are the relative ease and minimal trauma associated with implantation. The major disadvantages are that once in place, they become firmly attached to the bile duct wall and cannot be removed, and that tumor ingrowth through the mesh and overgrowth at the ends can occur. These stents, while quite helpful in the management of malignancies (especially in hilar obstructions), cannot yet claim to be the final solution to the problems associated with stent implantation in malignant disease . Nevertheless, the use of expandable metal stents has provided a most useful addition to the armamentarium of the interventional radiologist.
Surgical procedures involving the liver have undergone a significant revival, propelled by liver transplant techniques. Surgeons have taken a hard look at the liver and have found it to be less forbidding and more forgiving than was previously thought. Aggressive surgery has dramatically changed the outlook for patients who, only a few years ago, had been considered to have unresectable disease.
The major unresolved issue in 1995 is to define the role of PCD in a multidisciplinary context. Interventional radiologists do not work alone. They have to interact with other specialists, such as biliary surgeons and invasive endoscopists. Yet, the data that would back up any significant assertion about the role of PCD in this context are missing. There are very few good-quality, randomized comparative studies of PCD, and there are no efficacy studies or long-term studies on the impact of the technique on survival.
Important questions surrounding the use of PCD, such as the need for exchanging biliary catheters to prevent infection, are addressed in an empirical manner, with no controlled studies to back them. In her review article, Dr. Shapiro explores the important issue of whether all jaundiced patients actually need surgical, radiologic, or endoscopic decompression. However, her assertions about this issue are not supported by controlled data, as such important studies are lacking.
In a society with increasing constraints on health-care costs, it is important to analyze the cost-effectiveness of procedures. In the case of biliary decompression, this is not an easy task. While at a first glance both percutaneous and endoscopic techniques seem to be less expensive than surgical techniques, a longitudinal study may prove that, in some cases, they may be more expensive than even surgical bypass, if multiple admissions for sepsis and catheter exchanges are factored into the cost analysis. Moreover, the impact of the new techniques on quality of life, workdays lost, and other indirect costs must also be taken into account. Until such data are available, the controversies regarding unproven techniques will continue without resolution.