The present study was conducted at Saiseikai Chuwa Hospital, Sakurai, Nara, Japan. This hospital is located in a rural district in Nara, and the average age of patients in this area is very high. The records of the patients, who underwent ERCP in this facility from April, 2017 to April, 2018, were retrospectively analyzed. In total, 146 patients (232 procedures) underwent ERCP, patients aged <75 years were excluded, and 94 patients (158 procedures) were enrolled in this study. The patients were divided into the PS 0-2 group and PS 3-4 group. For both groups, clinical characteristics were analyzed, such as age, sex, PS, primary disease for ERCP, the prevalence rate of moderate or severe cholangitis [assessed by the severity assessment of acute cholangitis per the Tokyo Guidelines 13 (TG13) (8)], the use of antithrombotic drugs, history of gastrectomy and the Charlson comorbidity index (CCI) (9), which is a weighted index that assigns a different risk score (from 1 to 6) to 22 conditions and their related relative risk of 1-year mortality. Clinical outcomes were analyzed, such as performed procedures, the number of ERCPs for each patient, the technical success rate, adverse events caused by ERCPs, admission periods and fatal cases occurred within 30 days following ERCP.

In the facility, 5 gastroendoscopists performed ERCP. All ERCP procedures were performed by one operator and at least one assistant among 5 gastroendoscopists. A rule was established to complete the ERCP procedures within 1 h from scope insertion for all patients, as the ERCP procedure bears the risk of adverse events, such as pancreatitis, even when performed by endoscopic specialists, and it is estimated that a lengthy procedure bears the risk of adverse events following ERCP. According to the TG13 flowchart (10), not all procedures are performed in a single session for patients with moderate or severe cholangitis. In the first session, only endoscopic drainage was performed, followed, after the improvement of cholangitis, by other procedures such as endoscopic lithotomy. The TJF-260V (Olympus Medical Systems Co.) and ED-450XL8/B (FUJIFILM Co.) endoscopes were used, and for patients who had Billroth II or Roux-en Y reconstructions, the SIF-H290S (Olympus Medical Systems Co.) was used, which is a short-type single-balloon-assisted endoscope useful for ERCP in patients with altered gastrointestinal anatomy (11). For all procedures, carbon dioxide (CO₂) insufflation was used during ERCP to reduce post-procedural pain and abdominal distension (12,13). For the purposes of sedation during ERCP, midazolam and pentazocine were used, monitoring vital signs and the degree of sedation, all the while increasing the sedation if necessary.

The PS score (7) was labeled according to information on the general status prior to admission.

The severity of cholangitis was assessed by the severity assessment of acute cholangitis as per the Tokyo Guidelines 2013(8). The adverse events caused by ERCP procedures are defined as follows: i) Pancreatitis: The onset of new abdominal pain with at an least 3-fold elevation of serum amylase levels following ERCP; ii) hyperamylasemia: At least a 3-fold elevation of serum amylase levels following ERCP, without abdominal pain; and iii) biliary infection, including cholangitis, cholecystitis and liver abscess. The onset of the adverse events was evidenced after the ERCP procedures and did not exist before the ERCP procedures. Liver dysfunction is defined as the elevation of L-aspartate aminotransferase (AST) and L-alanine aminotransferase (ALT) that were at normal levels before the ERCP procedures and were found to be above normal levels after the ERCP procedures in the absence of any other cause. Mallory-Weiss syndrome is defined as the laceration of gastroesophageal junction caused by nausea or vomiting with or following ERCP procedures.

Comparisons between the study groups were performed with the Wilcoxon signed-rank test for quantitative variables and the Chi-square test for qualitative variables. A P-value <0.05 was considered to indicate a statistically significant difference. All statistical analyses were performed using IBM SPSS statistics version 22 (IBM Corp.).

The clinical characteristics of the patients in both groups are presented in Table I. A total of 94 patients (158 procedures) who were over 75 years old and underwent ERCP were divided into 2 groups as follows: The PS 0-2 group with 73 patients (116 procedures), and the PS 3-4 group with 21 patients (42 procedures). The average age did not differ significantly between the 2 groups. The ERCP rate in females was significantly higher in the PS 3-4 group than in the PS 0-2 group. The numbers of patients for each PS were 31 for PS 0, 36 for PS 1, 6 for PS 2, 4 for PS 3 and 17 for PS 4. Of note, the prevalence rates of moderate or severe cholangitis and the CCI were significantly higher in the PS 3-4 group than in the PS 0-2 group. The rate of anti-thrombotic drug use did not differ between the groups. In the PS 0-2 group, 9 patients had undergone gastrectomy (3 patients with Billroth I reconstruction, 3 patients with Billroth II reconstruction and 3 patients with Roux-en-Y reconstruction); however, in the PS 3-4 group, no patients had undergone gastrectomy. In both groups, no patients had undergone pancreaticoduodenectomy or choledochojejunostomy. The majority of the indications for ERCP in both groups were choledocholithiasis and malignant biliary obstruction (Table II).

The ERCP results for both groups are presented in Tables III-V. The performed procedures did not differ in either group. In particular, endoscopic biliary stent (EBS) and endoscopic lithotomy were frequently observed (Table III). Patients in the PS 3-4 group required frequent ERCP for one primary disease and longer admission periods (Table V). The reason was that patients in the PS 3-4 group had a higher prevalence rate of moderate or severe cholangitis (Table I). Only biliary drainage was performed for the first procedure, and other procedures such as endoscopic lithotomy were performed following the improvement of cholangitis. Moreover, a number of patients in the PS 3-4 group were often discharged to the nursing facility. Thus, several procedures and the adjustment of the discharge schedule required longer admission periods. Subsequently, the technical success rate for planned procedures was adequately high in both groups. There were no cases that required the suspension of ERCP procedures due to hypotension, bradycardia or respiratory depression caused by oversedation. In the PS 3-4 group, all planned procedures were completed (Table V). However, in the PS 0-2 group, 5 cases failed to achieve bile duct or pancreatic duct cannulation at the first ERCP. However, 3 cases were completely successful at the second ERCP, and 2 cases were changed to perform procedures under a non-endoscopic approach. One case after Roux-en-Y reconstruction failed to reach the papilla of Vater. Finally, the rate of adverse events caused by ERCP procedures did not differ significantly in both groups (Table IV). There were no severe adverse events, such as perforation and bleeding, which required endoscopic hemostasis or blood transfusion. Although 5 cases of post-ERCP-pancreatitis were observed only in the PS 0-2 group, they were improved only by conservative therapy. Despite the rate of adverse events by ERCP procedures not differing in both groups, there were 2 fatal cases during the 30 days following ERCP in the PS 3-4 group and no fatal cases in the PS0-2 group (Table V). The two fatal cases during 30 days following ERCP were the following.

i) Case A: An 86-year-old female with PS 4 underwent placement of a covered-type self-expandable metallic stent (SEMS) for biliary obstruction with pancreatic cancer. After 1 month of ERCP, she was admitted to the hospital for cholangitis due to SEMS obstruction. She then underwent SEMS cleaning and improved. However, she was readmitted for cholangitis due to a recurrence of SEMS obstruction (Fig. 1A). The prolapsed, obstructed SEMS (Fig. 1B) was removed and replaced by an EBS (Fig. 1C). Subsequently, although cholangitis was improved, her general condition gradually worsened due to anorexia caused by pancreatic cancer. As a result, she died 26 days after ERCP.

ii) Case B: A 98-year-old male with PS 4 underwent placement of EBS for biliary obstruction with pancreatic cancer (Fig. 2A). At 10 days after ERCP, cholangitis with EBS obstruction occurred; The EBS (across the papilla) was replaced with the plastic stent inside [i.e., above the papilla (Fig. 2B, arrowheads)], and the cholangitis gradually improved. However, at 10 days following stent replacement, he experienced a hospital-acquired influenza infection. Thus, his general condition worsened and he died 14 days after ERCP.