Today’s sleep-inducing medications reduce the risk of a patient not sleeping enough hours in the laboratory for a valid polysomnogram, but selecting when pharmacologic help is appropriate is not a simple decision.
Whether medicating patients undergoing a polysomnogram is appropriate is a question with a long history in the sleep medicine field. Pharmaceuticals that help patients fall asleep faster or stay asleep can ensure that technicians will be able to record enough hours of sleep for accurate testing. Capturing usable results from a complete sleep cycle on the first attempt means efficiency for physicians and less trouble for patients, who may be understandably reluctant to spend another night in a laboratory bed with elastic bands around their chests and abdomens, clips on their finger, and various electrodes on their heads.
However, while the benefits are obvious, so are the down sides. Any medication designed to either speed sleep latency or, conversely, reduce drowsiness can affect the results of a sleep study. Therefore, physicians must carefully weigh the benefit of faster, surer sleep against the risk of compromising their study results. While newer sleep-inducing medications with short half-lives that, in theory, should impact only the first few hours of sleep in a study are now available, physicians contacted by Sleep Review say it is still a highly individual decision-making process. It is paramount to consider the patient before considering medication, they say.
A Patient-centered Approach
Physicians did not always factor a patient’s unique circumstances into the equation, says Daniel O. Lee, MD, medical director for the Pitt County Memorial Hospital Sleep Center, in Greenville, NC, where around 4,800 sleep studies are done annually. When he began conducting sleep studies in 1989, sleep physicians would forbid all patients from taking medications before a sleep study because they believed that a patient who was under the influence of a prescription drug would not provide the best samples of his or her own, unique sleep architecture. Sometimes this approach was successful. Other times, the patients were unable to sleep and would stare wide-eyed at the sleep technicians and physicians who stared back at them, hoping for sleep to overcome the patients so the proper data could be collected.
In addition, this blanket approach was difficult for patients who had relied on a prescription drug to get to sleep for years. In preparation for their polysomnograms, they were forbidden from taking their usual night-time sleeping aid for as long as 2 weeks leading up to a study. After days or weeks of restless nights, they would arrive for their sleep tests in a state far from what they considered to be their normal sleep-wake schedules.
Because some patients were not falling asleep, the results that the physicians did receive were often inconclusive. So, in the mid ’90s, a different philosophy began to take hold, Lee says. The paradigm shift focused on the individual patient and his or her specific routine and suspected sleep disorder.
“It all depends on the individual,” Lee says. “A lot of times we try to make it one simple formula, but unfortunately the human being is a lot more complex. We have to use clinical judgment to find what we are looking for.”
If a patient is being tested for excessive daytime sleepiness, for example, no medication should be given because it would interfere with the sleep architecture and could make interpretation of the patient’s symptoms difficult: Is the daytime sleepiness caused by the medication or a sleep disorder? However, for patients who have insomnia and a possible secondary disorder, such as obstructive sleep apnea, prescribing a low dose of a sleep-inducing medication may be the only way for physicians to be able to get these patients to sleep long enough in the laboratory for the secondary disorders to be diagnosed.
“To treat a sleep study in isolation with guidelines governing every single thing you do, I think that just removes it from clinical relevance,” says Scott Eveloff, MD, ABSM. “There are some things that still have to be on a case-by-case basis, even in this age of evidence-based medicine and computerized medicine and guidelines. Call me old-fashioned, but patients deserve at least a case-by-case consideration of clinically personal aspects.”
At Somnitech Inc, Overland Park, Kan, where Eveloff works, physicians do not use across-the-board recommendations on medicating sleep study patients, even though they conduct a large volume of studies. He believes that a predetermined approach to medicating patients would yield unnecessary false positives, as well as false negatives, on tests. Instead, Eveloff keeps a number of questions in mind before he will prescribe any medication to help a sleep study patient fall asleep in the laboratory. These include:
• What is the effect of the medication on sleep architecture and thus results?
• What are the effects of the medication on breathing during sleep? (Will it create symptoms of a sleep disordered breathing disease that was never there in the first place?)
• What is going to give the more invalid result: interruption of sleep due to a first-night effect in the sleep laboratory or a medication that could possibly change the sleep pattern?
To make his final decision, Eveloff factors his answers to these questions into an individualized risk-benefit analysis. This is time-intensive and includes either interviewing his patients directly or thoroughly reviewing the information provided by the referring physician. In some cases, when a referred patient is anxious about the study or is a complex case, Eveloff might even do both.
In most cases, it is unnecessary to use medication to help a patient achieve enough hours of sleep in the laboratory for a proper study. However, the option needs to be available, say those who favor a patient-centered approach. Eveloff says that he has medicated only about 5% of his patients in the past and uses zolpidem exclusively because he feels that the half-life of the drug is acceptable.
Early research backs him up. A study in The Journal of Clinical Sleep Medicine1 found that pretreatment with zolpidem significantly improved polysomnographic quality and decreased the need to repeat tests. Patients that did not get zolpidem had a higher incidence of uninterpretable studies. Sleep efficiency and sleep latency also increased in medicated patients.
Eveloff says the authors of the study adopted a liberal view of prescribing zolpidem after analyzing the findings, but he has reservations. “The study did not address whether, in people who turned out to have sleep apnea, if they would have had it if they had not been given zolpidem,” he says. “It opens the door toward a prospective study, and I personally think that is what is needed.”
Eveloff will typically only prescribe zolpidem to patients who did not achieve an adequate degree of sleep for a proper diagnosis on the first test. Although reluctant to prescribe zolpidem for the initial test, he considers it for patients if there is a high suspicion of sleep apnea. He is also more comfortable in considering prescriptions for people who are convinced they will be unable to sleep in the laboratory.
Holding a conservative view of when and how to use sleep-inducing medications in sleep testing does have disadvantages. For one, it can mean more after-hours calls to the physician from the sleep laboratory and it increases the possibility of having interrupted studies that must be redone, which means greater cost. Eveloff says that redoing a study can run from between $800 and $1,000.
Finding the Balance
“I still think that patient disease process and patient evaluation come first,” Eveloff says. “Financial and monetary considerations come second. I feel very reluctant to up-front just give people [zolpidem] out of a concern that you are going to reduce the number of repeat studies. I, personally, am nowhere near that point yet.”
Whether to allow use of medication in a sleep study can also depend on what patients are accustomed to. The sleep clinic Sleep Insights in Rochester, NY, tries to mimic patients’ home environments as closely as possible, and that includes inviting patients who regularly take prescription medications to help them sleep to bring those pharmaceuticals with them.
Ken Plotkin, MD, president and medical director of Sleep Insights, says that he would rather interpret polysomnogram results from patients that had a reasonable night’s sleep with the help of a drug they take regularly than work with results from patients that had overly fragmented, minimal sleep because their usual sleeping aid was taken away.
However, prescribing sleep-inducing drugs to people who are not currently taking such medications is a different matter. Plotkin generally does not do so, even when the patient thinks he or she may have a problem falling asleep in the laboratory. “If they use medicine at home to get to sleep, they come to the lab and get to sleep [with the help of the drug],” he says. “If they do not take medicine at home and have trouble getting to sleep, they will come to the lab and have trouble getting to sleep, but eventually get to sleep.”
He also believes that allowing patients to bring in medications they use regularly for the sleep test sometimes may yield a more relevant study result because the situation then more closely resembles the patients’ regular sleeping environment. But the effectiveness of drugs taken during a sleep study is a double-edged sword. Some medications can accentuate obstructive breathing and snoring. Other medicines, such as dopamine-mimicking drugs, suppress periodic limb movements or the arousals and the brief awakenings of limb movements.
There are no simple answers to the question of when to use sleep-inducing medications in a sleep study. For Lee, Eveloff, Plotkin, and sleep physicians like them, the best approach right now is the individualized one.
Stephen Krcmar is a contributing writer for Sleep Review.
1. Letttieri CJ, Eliasson AH, Andrada T, Khramtsov A, Kristo DA. Does zolpidem enhance the yield of polysomnography? J Clin Sleep Med. 2005;1:129-131.