Dear Doctor: I've seen a lot of dogs recently in grocery stores and restaurants wearing those yellow service dog vests, but some of them can't obey even simple commands like "sit" and "stay." They seem to really be pets. What do trained service dogs do? Whom do they help?

Dear Reader: It's a shame when dog owners misrepresent their pets as service animals. No doubt some are legitimate "emotional support" animals, necessary companions for their owners to be able to spend time in public spaces. But federal law states that a service dog is one that has been specially trained to physically assist a person with a disability, including -- and we're quoting the law here -- "a physical, sensory, psychiatric, intellectual or other mental disability."

It is on that basis that these devoted animals are granted access to venues and areas not open to other pets. With their rigorous training and specialized skills, service dogs open up the world for their handlers, and keep them safe within it.

Service dogs perform hundreds of tasks for more than a dozen types of disabilities. We're all familiar with guide dogs, which help people with impaired vision. They lead their handlers around obstacles like a park bench, a low-hanging awning or a hole in the ground. They warn them of changes in elevation, like a curb or the edge of a subway platform. They can follow a designated person, like a waiter in a restaurant, or find their handler an empty seat in a public space. And though their handlers are the decision-makers in the partnership, guide dogs have been taught "intelligent disobedience." When given a command to walk forward, if danger is present, like a sudden drop-off or oncoming traffic, they will refuse.

For people with impaired hearing, specially trained dogs become their ears. With a touch of their nose or a gentle paw, they can signal a ringing telephone, a crying baby, a smoke alarm, an alarm clock, a family member calling the handler's name, computer beeps, cellphone alerts and a person's arrival.

People with physical disabilities or missing limbs rely on their service dogs to help with mobility. These dogs can pull a lightweight wheelchair, offer assistance by bracing their handlers as they get up or down, and help their handlers rise if they should fall down. They can open doors, turn light switches on or off, and pick up objects as small as a dime.

Seizure dogs, which are trained to recognize their handlers' physical symptoms, can summon help by calling 911 via a special life-alert system, or provide physical stimulation. Like many service dogs, they are trained to retrieve medication. Diabetic alert dogs use their sense of smell to detect episodes of high or low blood sugar and warn their owners. Severe allergy alert dogs let their handlers know about life-threatening allergens nearby.

Service dogs are remarkable in their training and dedication. And though it's tempting to give them a pat or say hello, please don't. Service dogs out in public are at work. Correct etiquette is to ignore them, so they are not distracted from their job.

(Send your questions to askthedoctors@mednet.ucla.edu, or write: Ask the Doctors, c/o Media Relations, UCLA Health, 924 Westwood Blvd., Suite 350, Los Angeles, CA, 90095. Owing to the volume of mail, personal replies cannot be provided.)

Dear Doctor: Do you, or do you not, need to take a whole course of antibiotics even when you begin to feel better? I've always dutifully done so, but new reports seem to suggest otherwise.

Dear Reader: Antibiotics have undoubtedly saved countless lives. My grandfather died from pneumonia in the 1940s, before antibiotic use became common, and I often wonder about his likelihood of survival in the 1950s, when the drugs were more readily available. Today, we've almost come full circle. We now live in a world in which some bacteria no longer succumb to the common drugs used against them, and we fear that the overuse of antibiotics may lead to further resistance. At the same time, more powerful antibiotics are used only when an infection is resistant to all other antibiotics, and drug companies lack incentives to develop new antibiotics.

That brings us to the essential question: How best to prevent drug resistance among patients given antibiotics? In July, a group of infectious disease specialists and microbiologists in England published an editorial in the British Medical Journal arguing that lengthy courses of antibiotics could actually increase resistance. Continuing to give antibiotics after they've done their job doesn't make sense, they said, and isn't backed up by science. They have a point. Although a few studies have assessed the appropriate duration of antibiotic use for specific conditions, overall, research is slim.

The "take all your antibiotics" advice stems from a long-held worry that bacteria not killed by the antibiotic could develop resistance to that antibiotic. The problem is, bacteria are not one-size-fits-all. Some bacteria, such as those that cause tuberculosis, gonorrhea, salmonella and malaria, can quickly become resistant to antibiotics, so it is important to kill them completely. Other bacteria are slower to develop resistance, so to have some lingering bacteria is less problematic. What is more worrisome about long regimens of antibiotics is that they also affect other bacteria that reside normally in our bodies, not causing infections. Not only do the "good" bacteria suffer, some of the "bad" bacteria can become resistant to that antibiotic and later cause infections that cannot be killed by the prior drug.

As I mentioned, some studies have indeed assessed the duration of antibiotic regimens -- and the results have been mixed. For kidney infections, a seven- to 10-day course of antibiotics has shown equal efficacy as a 14-day course. However, in other circumstances, such as ear infections in children, there is greater benefit in taking antibiotics for 10 days than five days. This also is the case for strep throat.

Complicating matters is the nature of various antibiotics themselves. The antibiotic azithromycin, for example, has a half-life of three days, meaning that although a typical course is only five days, the drug stays in your system for many days after you finish.

In summary, we obviously need more studies on the duration of antibiotic regimens. But the authors of the recent editorial skipped over the major cause of the antibiotic-resistance problem: the over-prescribing of antibiotics. The federal Centers for Disease Control and Prevention has estimated that 30 percent of all antibiotic prescriptions are unnecessary. Most of this overprescribing is for upper respiratory symptoms.

We should focus less on the duration of the antibiotic regimen and more on whether the antibiotic is necessary. If patients and medical practitioners use antibiotics judiciously, we may be able to curb the rising rate of antibiotic resistance.

(Send your questions to askthedoctors@mednet.ucla.edu, or write: Ask the Doctors, c/o Media Relations, UCLA Health, 924 Westwood Blvd., Suite 350, Los Angeles, CA, 90095. Owing to the volume of mail, personal replies cannot be provided.)

Hello again, dear readers. We hope you've been having an enjoyable summer and that our recent columns about sunscreen, ticks and mosquitos have come in handy. We've been hearing from a lot of you and, as always, you've given us great information to share, as well as a few things to clarify.

-- Let's start with the column about floaters. Those gently drifting specks and strands some of us get in our field of vision generated quite a few letters. Some were sweet thank-you notes, to which we say, thank you for reading the column! And a few disagreed with the idea that, over time, one's brain will adjust and "ignore" floaters.

A reader from Baker, Louisiana, put it most succinctly: "Every time I read an article or hear a doctor say that given time, 'your brain will tag floaters as unimportant information and, for the most part, will let you forget them,' I have to laugh. Laugh through gritted teeth, to be sure, but laugh nonetheless."

Some of you wrote with questions regarding the newest research into the use of lasers to reduce or remove floaters. It's an intriguing area of study, and we will look into it for you in the future.

In that same column, we stated that a sudden increase in floaters, when accompanied with flashes of light, could indicate a serious problem. In answer to the reader who (very politely) took us to task for not specifying what that problem might be, the answer is that it may indicate a retinal detachment. That's why, as we said, that set of symptoms means it's very important to check in with your eye doctor.

-- After publication of our column about yellow fever, we heard from many world travelers who want to know whether they need to get re-vaccinated. We have good news. According to guidelines updated by the World Health Organization in 2013, a single primary dose of the yellow fever vaccine gives lifelong protection. A booster dose of the vaccine is not needed. The Centers for Disease Control and Prevention have adopted the WHO recommendations.

-- After our column about testing for prostate cancer, a reader asked for information about the IsoPSA test. This is a new test and it is now undergoing additional study. Initial findings suggest that the IsoPSA may be more precise than the existing prostate antigen test at distinguishing cancer from benign conditions, and in identifying patients who are at higher risk of developing the disease. We'll keep an eye on future developments.

-- Finally, thank you to all of you astronomers who let us know that the most recent total solar eclipse in North America was actually more recent than what our column stated. An excerpt from one of the letters:

"There was a total eclipse in Canada in February of 1979 and I know that because I was visiting my parents in Winnipeg, Manitoba, at the time and was very blessed to witness the total eclipse at approximately 9:30 a.m. It was an amazing sight, corona and all."

As ever, thank you for taking the time to write to us. We know we are fortunate to have readers like you.

(Send your questions to askthedoctors@mednet.ucla.edu, or write: Ask the Doctors, c/o Media Relations, UCLA Health, 924 Westwood Blvd., Suite 350, Los Angeles, CA, 90095. Owing to the volume of mail, personal replies cannot be provided.)