The incidence and risks of liver biopsy in non‐cirrhotic patients: An evaluation of 3806 biopsies

Liver biopsy plays a crucial role in the diagnosis and management of liver diseases. For the past decade, this invasive procedure has become a safe one with the prevailing application of an ultrasound‐guided method, the use of thinner gauge needles and improved operational techniques. Over the past years, the debatable issue of liver biopsy has mainly focused on the safety and suitability of a shorter observation time with respect to cost savings.¹,²,³ Referring to this issue, Beddy et al even shortened their observation time to 1 hour and indicated that only one haemorrhagic complication occurred within one hour amongst 500 liver biopsy occasions. There were no recorded delayed complications or deaths at follow up.⁴

Generally, we agree with the conclusion that outpatient liver biopsy is safe when done in a setting that provides close observation for 1 hour after a biopsy. However, we keep a conservative and different view on the sufficiency and safety of 1 hour observation time after a liver biopsy. We have performed a total of 3806 liver biopsies using 18‐gauge needle on 2980 patients (males 1817, females 1163, mean age was 41.3 years, range 17–72 years), performed in a medical centre and in a core regional hospital in southern Taiwan from Jan 1996 through Sep 2006. Our outpatient biopsy was assigned to a 6‐hour observation time. The reasons for the liver biopsy studies on these patients, were either a result of chronic viral hepatitis or a surveillance of abnormal liver function tests. There was no clinical evidence of cirrhosis diagnosed by high‐resolution ultrasound before a biopsy. Patients who have thrombocytopenia or the presence of oesophageal varices, ascites, splenomegaly, signs of portal hypertension, or encephalopathy by other imaging studies as well as bleeding tendencies, were excluded. These procedures have been done by six well‐trained, board certified hepatologists. Apart from minor complications such as pain, nausea, and vomiting, a total of 12 patients (0.32%) suffered from haemorrhagic complications 4 to 12 hours after the biopsy (table 1). Our results were comparable with those of Beddy et al in terms of major complications and mortality. Although there was no death at follow up of one week, four patients did experienced hypovolemic shock. In one patient, haemobilia, presenting with insidious onset of tarry stool, obstructive jaundice, and hypovolemic shock, developed 24 hours after the biopsy. The total in‐hospital days therefore were extended and ranged from 1 to 7 days. In rare occasions haemorrhagic complications such as haemoperitoneum, haemothorax, and haemobilia did occur with potential risk of morbidity and mortality after the procedure. Noteworthy is that those with poor performance such as liver cirrhosis and bleeding tendency were excluded from our study. Major complications occurred in four patients after the regular observation time. This does, indicate a possible need of a longer time of observation. Therefore, our study with a 6‐hour observation time basis concluded that although ultrasound‐guided liver biopsy is generally a safe procedure, major haemorrhage complications may occur more than one hour after liver biopsy. A longer time of observation may be more suitable with this potential risky procedure.

Table 1 The characteristics of 12 patients with major complications after liver biopsy.

Patient no.	Sex/age	Complications	Time of occurrence (hrs)	In‐hospital days
1	M/35	Haemoperitoneum	6	3
2	M/53	Haemoperitoneum	6	3
3	F/55	Haemothorax	12	1
4	M/52	Haemothorax with empyema	6	7
5	M/53	Massive subcapsular haematoma with shock	12	1
6	M/35	Haemoperitoneum	12	1
7	F/60	Haemoperitoneum with shock	6	7
8	F/53	Haemoperitoneum	5	5
9	F/55	Haemoperitoneum	5	5
10	F/42	Haemoperitoneum, haemobilia, shock, and transient obstructive jaundice	24	7
11	M/65	Haemoperitoneum with shock	6	5
12	F/53	Haemoperitoneum	4	3