Hydroxychloroquine

What this drug is for

Hydroxychloroquine is a tablet taken every day. It started life as a malaria drug and it's now one of the standard long-term treatments in lupus, used in some other rheumatic diseases too. It doesn't work like a painkiller at all, because it works slowly, over months rather than hours.

Most drug pages would start with what a drug does to symptoms, and this one starts somewhere else, because the strongest evidence here isn't about how you feel. It's about how long people live. Two large studies point in that direction, and neither assigned anybody to take the drug, so neither can prove it caused what they found.

Pregnancy is the question our sources answer worst, and the hole in them is not a small one. One study looked at 399 pregnancies in 324 women with lupus and asked a single question, which was whether the drug cut pre-eclampsia, and it consistently found no effect. That is one outcome rather than a safety assessment, so pregnancy advice here needs your own team and your own antibodies.

The survival figures

One study followed 2446 people with lupus in China, where there were 221 deaths over the follow-up. People taking this drug had about half the death rate of people who weren't, at a hazard ratio of 0.49, with the true value somewhere from 0.35 to 0.67. That's a large study reporting a large difference.

A second study used national health records in Taiwan. It matched 2287 people on the drug against 2287 who weren't, then followed both groups for an average of 7.6 years. Deaths ran 7.4 percent in the treated group against 10.8 percent, at a hazard ratio of 0.68, from 0.56 to 0.82.

Read those carefully, because they watched what happened and tested nothing. People who get prescribed a drug and keep taking it differ from people who don't, in ways an analysis can't fully adjust away. That said, two large studies in two countries pointing in one direction is the strongest signal this site holds for any drug in lupus.

The eye risk, and why it's handled the way it is

This drug can damage the back of the eye, which is its one serious long-term risk and the reason your eyes get checked every year. Three things about it count for more than the headline number does. Each one changes what you do about it.

First, it builds with time rather than arriving all at once. At the recommended dose the risk is under 1 percent up to 5 years, under 2 percent up to 10 years, and almost 20 percent after 20 years. A separate cohort of 537 people found 4.3 percent overall, at 1 percent in the first 5 years and 11.5 percent between 16 and 20 years.

Second, it's usually silent while it happens. One review describes the damage as often subclinical, generally irreversible, and able to keep progressing even after the drug stops, so waiting until your vision changes is waiting too long. That's why the check is a scan rather than a question about how you're seeing.

Third, catching it early changes what happens next. Mild damage is unlikely to get worse once the drug stops, while severe damage may keep worsening for years afterwards. Those are two different futures, and screening is what is between them.

The dose, and why your weight is in it

The guideline sets a ceiling on the daily dose and works that ceiling out from your real body weight rather than your ideal weight, because real weight tracks the eye risk better. This page prints no figure. Your dose is a calculation for your own team, and it depends on your kidneys as well as on your weight.

What's worth knowing is that the ceiling exists and that people do end up over it. One clinic audited 427 of its own patients and found that after allowing for weight and kidney function, 10 percent were prescribed above the limit, which was 31 of 312. So asking whether your dose is under the ceiling is a reasonable question to put to your team.

That audit found something else worth repeating. Formal documentation that the eye risk had been explained appeared in the notes of only one third of patients. So if it hasn't been talked through with you, that's common rather than unusual, and it's a fair thing to raise.

What raises the risk

Dose and duration come first, and they're the two the guideline leans on. Beyond them it names kidney disease, taking tamoxifen, and starting the drug at an older age. Smaller studies add more, and those should be held loosely.

One study of 92 people in Taiwan found short sight was associated with retinopathy at an odds ratio of 5.03, with a true value anywhere from 1.29 to 24.79. That range is enormous, which is what 92 people buys you in a study. It also found lower body weight was associated with it.

One study of 345 people found two things associated with damage appearing early, meaning within the first 5 years. Lupus anticoagulant positivity came in at 4.2, from 1.2 to 15.5, and high cholesterol came in at 5.6, from 1.5 to 21.5. Only 10 people in that study had early damage, so both of those figures rest on a very small group indeed.

The number that should change how you think about this

One study measured the drug in the blood of 108 people who all said they were taking it. Some 41 percent came in under 500 nanograms per milliliter, which is the level used to mean somebody isn't really taking it, and of those, 19 percent had nothing that could be detected at all. That's a striking result for a drug people believe they are taking.

Higher disease activity correlated with the low levels, at a P value of 0.003, and that held after allowing for depression, at 0.0007. Low levels accounted for 32 percent of the variation in disease activity scores across the group. So a third of the difference in how active people's lupus was correlated with how much drug was in them.

That study looked at one moment in time, so it can't say whether missed doses drove the disease activity or active disease made people stop taking it. What it does show is that a drug this quiet is easy to drift away from without noticing. A national study in New Zealand found 54.5 percent of people collecting it had high adherence, meaning they held a supply at least 80 percent of the time. If a treatment seems not to be working, that's a question worth asking before anything gets changed.

Common misconceptions.

Related terms.

• Lupus, the condition most of this evidence comes from.
• Methotrexate, the other long-term tablet used across these diseases.
• Lupus and blood tests, for what the monitoring bloods are looking for.