27 Jun 2013


Not surprisingly, AL and K passed the College examination with flying colors. Shortly before the examination, they asked GW and me to conduct a mock examination for them. Personally I think such kind of exercise is quite useless. But as Prof RK said, when many useless things add up together, they may turn out useful. To go along this line, I asked a useless question that would never appear in the examination.

“A 38-year old man learned from the news that a rich man in town is dying from liver failure. He offers to donate part of his liver to him, but demands a compensation of $500,000 in return, which the family for sure is willing to pay. He argues that he owns his body and should have autonomy over its use. Besides, he believes that his request is reasonable given the sacrifice he is going to make. As a doctor, do you agree with his points?”

After thinking for a while, the candidate replied, “This is certainly not acceptable. Above all, doing no harm is the fundamental standard of medicine. We cannot risk his life to save a richer person.”

“Wait a second,” I interrupted. “Do you mean you are against living donor organ transplantation?”

Seeing it could not be right, the candidate changed strategy, “Not really. We also need to see if the patient had problems with alcohol and the reason for liver failure.”

“He does not drink alcohol. He had decompensated hepatitis B-related cirrhosis.”

“We should see whether the patient really requires liver transplantation and is fit for the procedure. Besides, we shouldn’t let someone jump the queue just because he can pay.”

“All right,” I said. “In other words, if the patient is at the top of the waiting list but a cadaveric liver is not available yet, you would find the proposal acceptable. In fact, you may argue that such a move would even shorten the waiting list for other patients.”

“No, no, no. Organ trading is illegal in Hong Kong.”

“That is true,” I said, “but what is the ethical basis of the policy?”

“I give up.”

At least we learned one thing that afternoon. Principles are not subject to conditions. There is no if or when. You cannot say the government should not run students over with tanks if they are protesting peacefully. They just shouldn’t.

20 Jun 2013


We held the professional examination last month. After examining the first five students, Professor RY remarked on the standard of the class at the coffee break. Many of us were astonished. It took us at least several months to arrive at the same conclusion. “Of course this may be sampling bias. I will examine the rest of the students first,” he said when the next batch of students entered.

When I told my friends the story, they were not impressed. “He doesn’t pay attention during examinations,” HC said. “He even dozed at the last college exam.” I did not argue. However, after so many years, I am sure he does not need ten minutes to decide who is a competent doctor.

Once, a famous director was invited to the rehearsal of a play. A moment later, he fell asleep. The cast was very embarrassed but had to continue. After the first act, the producer could not tolerate any longer and woke the director up. “We are sincerely asking for your comments, and you simply dozed off!”

The director rubbed his eyes and mumbled, “That was my comment.”

13 Jun 2013


I gave a lecture at the refresher course last Monday. Two days later, I received a list of questions from a student. I have never answered questions in my blog before, but then I have never received so many questions from a student all at once either. The questions are practical and insightful and should be shared with other students.

1. I was told that in liver cirrhosis progression, it usually starts with deranged clotting profile, then albumin, then bilirubin, then ascites and finally encephalopathy. Is that true that when patient developed, let's say, ascites or encephalopathy, it indicates a more severe cirrhosis?

Answer: The sequence of laboratory changes is not absolute. However, it is true that clinical complications such as ascites and encephalopathy usually indicate more severe disease and are associated with poor prognosis.

2. In monitoring patients with cirrhosis, do we monitor by Child-Pugh grading + HCC surveillance + varices screening + hepatorenal syndrome screening? Do we need to repeat ultrasound for progression of features?

Answer: You have included different concepts here: assessment of the severity of cirrhosis and complications screening. Currently, most doctors use the Child-Pugh score or the model for end-stage liver disease (MELD) score to assess the severity of cirrhosis. The purpose is to predict the prognosis and prioritize patients for liver transplantation.

HCC surveillance is performed to detect early HCC so that patients may receive curative treatment. That said, this should only be done in patients who are candidates for curative treatment. For example, a patient with Child’s C cirrhosis and multiple comorbid illnesses cannot undergo liver resection or locoregional therapy even if a small HCC is detected. Surveillance would not be helpful.

Screening for varices is also recommended to reduce the risk of variceal bleeding.

Finally, the main role of ultrasound is for HCC surveillance. We use other parameters to assess the severity of cirrhosis.

3. Once patient is diagnosed to have liver cirrhosis, when should we start screening for varices? and by what mean? Ultrasound or endoscopy?

Answer: Unless the patient is moribund, all cirrhotic patients should be screened for varices by OGD.

4. When we say ultrasound imaging to look for presence of varices, do we mean hepatic doppler ultrasound that the flow in different hepatic vasculature may suggest the presence of esophageal varices?

Answer: In good hands, splenic varices may be visualized by ultrasound. While this feature confirms the presence of portal hypertension, it cannot replace OGD. It is bleeding from esophageal or gastric varices that we want to prevent.

5. For HCC surveillance, do you mean serum AFP + LFT + CT (or USG?) scan regularly?

Answer: Good question. For some hepatologists, this means 6-monthly USG and AFP testing. Liver function test does not detect HCC. CT is more accurate but involves radiation and has not been tested in the screening setting. Triphasic CT is usually reserved for confirming the diagnosis of HCC when a liver nodule is identified by USG. Interestingly, the current American guidelines only suggest USG surveillance and discourage the use of AFP. This is highly debatable.

6. In bleeding esophageal varices, do we need to give both octreotide and terlipressin together? What is the usual dosage given and are they given as IV infusion or bolus?

Answer: Either somatostatin analog (e.g. octreotide) or vasopressin analog (e.g. terlipressin) would do. The dosage we are using is octreotide 50 mcg iv stat, followed by 50 mcg/h infusion; and terlipressin 2 mg Q4-6H iv. There have also been studies showing that terlipressin infusion may be better than bolus injections.

7. In variceal bleeding, do we need to correct the deranged clotting profile by transfusing fresh frozen plasma? and also correct any platelet derangement?

Answer: Many doctors do so, but this practice is not adequately tested.

8. Do we need to prophylactically give lactulose to all cirrhotic patients with variceal bleeding to prevent development of hepatic encephalopathy? or do we give only when patients develop symptoms and signs of HE?

Answer: Because patients with variceal bleeding are at high risk of developing hepatic encephalopathy and lactulose carries few side effects, we usually give it prophylactically.

9. In secondary prophylaxis of variceal bleeding, you mentioned an option of endoscopic variceal ligation, but I thought you have already banded all varices during last variceal bleeding? or do you mean those new varices developed? Also, do we need to start beta blocker +/- nitrate immediately after first episode of variceal bleeding?

Answer: It takes an average of 5 sessions to eradicate all esophageal varices. A clinical trial from Spain showed that pharmacological therapy is more effective than endoscopic variceal ligation for secondary prevention of variceal bleeding and result in less complications (N Engl J Med 2001;345:647-55). However, a subsequent meta-analysis showed that combining endoscopic variceal ligation and pharmacological therapy would further reduce the chance of rebleeding (Aliment Pharmacol Ther 2012;35:1155-65). At our center, we do both.

Immediately after variceal bleeding, the patient should still be on vasoactive drugs such as octreotide or terlipressin. You do not need to start beta-blockers yet. Besides, it is a bad idea to give a drug that would lower the blood pressure right after active bleeding. Instead, beta-blockers and/or nitrates may be started after the acute episode settles.

10. If we are to give lactulose to patient during variceal bleeding, since patients are kept NPO, can he still tolerate an oral lactulose? How about in the case when patient is drowsy and confused as in grade 2-3 encephalopathy or patient in hemorrhagic shock, is it still safe to give oral lactulose?

Answer: We can keep the patient nil by mouth except medications. If the patient has aspiration risk, medications should be delivered via nasogastric tube instead.

11. In patients with ascites with deranged renal function (creatinine >133), do we routinely stop diuretics or paracentesis for 2 days to assess hepatorenal syndrome?
And even with albumin infusion, ascites will still come back right? And is this because hypoalbuminemia is not related to pathophysiology of ascites? And in this case, do we just leave the ascites untreated and observe?

Answer: It is important to stop diuretics and paracentesis in cirrhotic patients with acute kidney injury. They may well be the culprit.

Albumin infusion is for the prevention of paracentesis-induced circulatory dysfunction, not ascites. Therapeutic paracentesis relieves ascites rapidly but does not alter the underlying pathophysiology of portal hypertension and salt retention. Therefore, ascites will recur after paracentesis unless the patient is well controlled with salt restriction and diuretics. However, some patients will still have refractory ascites and require either repeated paracentesis or TIPS.

12. I'm sorry to complicate the case further. What if the patient has co-morbid cardiovascular condition (eg. hypertension, heart failure) that requires the use of diuretics? Do we still stop it for assessment? If patient is on ACEI, do we need to stop it for HRS assessment?

Answer: You have to assess the fluid status. If the patient is having fluid overload, you do not have much choice. For your last question, cirrhotic patients are very sensitive to ACEI because renal perfusion is often suboptimal. In case of renal deterioration, we have low threshold to stop ACEI.

6 Jun 2013


Curiously, this is the first time I wrote on my blog on our anniversary. This year is also special in that my brother Roy is getting married later this month.

Roy reminds me much of Levin in Anna Karenina. He expressed how pointless the wedding preparation and ceremony were, but ended up doing everything others did. I wish he would enjoy it as much as Levin did in the end.

Every Christmas, our parish priest tells us a story about love. Every Christmas, he tells the same story.

Once there was a poor couple. The husband had a watch, but it had no straps. One Christmas, the wife gave him a pair of leather straps as present. The man was very surprised, “How did you get the money?” The wife replied that she sold her hair. The man then gave her a present. It was a beautiful comb. This time, it was the wife who was surprised. “I sold my watch,” the man explained.

At this point, Angelina commented, “You can still use the comb with short hair. Besides, the hair can grow back. The man was less lucky.” I replied, “I think he was very lucky. The present was love.”