Introduction
It is important to examine the reliability of test buttons on smoke detectors in the light of a two-year investigation (1994-1995) of smoke detectors in U.S. homes. The federal government, in two recent surveys, sought answers to the following questions, among others:

• How many smoke detectors are there in the nation's homes?
• What types are they?
• How many are in working order?
• How many are not?
• Why not?

This article will deal with one important finding of these studies, namely, questions involving the test button. In addition to the reliability question, it is necessary to address certain practical problems that make it difficult, even impossible, to test smoke detectors by pressing the test button.

A Court Room Test
Perhaps a good way of getting some visibility on the reliability question is to offer the following court room scenario of an attorney interrogating the CEO of a smoke detector manufacturer. In this case, the attorney is serving as counsel for the estate of the Smith family, all of whose members died during a home fire. The CEO's company, ZEUS, manufactured the smoke detectors installed in the Smith household.

ATTORNEY: Please identify yourself for the record.

CEO: I am Harold Jones, Chief Executive Officer, ZEUS Company.

ATTORNEY: Were there ZEUS smoke detectors on the premises of the Smith home?

CEO: Yes.

ATTORNEY: Did these smoke detectors give the Smith family the "early warning" regarding a fire that your company claims?

CEO: Yes, they're designed to do exactly that.

ATTORNEY: Why, then, did all members of the family perish? Do you want to stand by your previous answer?

CEO: Well, I don't know for a fact that the detectors did go into alarm.

ATTORNEY: So you don't really know whether any of your smoke detectors will sound the alarm in the event of a fire?

CEO: On the contrary, we instruct our customers to test their smoke detectors frequently.

ATTORNEY: How are these instructions communicated to the customer?

CEO: Each smoke detector package includes a user's manual which tells the customer to test the smoke detector.

ATTORNEY: Please read for the record the section dealing with testing.

CEO: (reading) "For a complete weekly test of your smoke detector, firmly press the test button on the cover for a few seconds. The smoke detector will then make a loud, continuous beeping sound if it is operating properly."

ATTORNEY: With the court's permission, we introduce as evidence the document just read by Mr. Jones.
ATTORNEY: (Then, turning to the CEO) Do you recommend any other test?

CEO: No.

ATTORNEY: Let's be sure we understand you. You're saying there's only one acceptable test for making sure the smoke detector works, and that's by using the test button. Is that correct?

CEO: Yes.

ATTORNEY: With the court's permission, I hereby introduce into evidence a report entitled, "Fire Incident Study - National Smoke Detector Project:, U.S. Consumer Product Safety Commission, January 1995. I read from that a section of that report (on page 14) dealing with "Detectors That Should Have Alarmed". "Eight percent [of the detectors that should have alarmed, but did not] were clogged with dust/dirt, and five per cent showed signs of insect infestation."

ATTORNEY: (Continuing with the witness) Mr. Jones, please answer this question: Will smoke enter the sensing chamber under the conditions just cited, that is, those with the clogging dust/dirt and insect infestation?

CEO: Probably not.

ATTORNEY: You can be more positive than "probably", can you not?

CEO: Yes, smoke would not enter the sensing chamber of detectors whose sensors were clogged with dust, dirt or insects.

ATTORNEY: Please listen carefully to this statement, Mr. Jones. That means, then, that even though pressing the test button causes the alarm to sound, it still doesn't mean that the alarm would sound as a result of smoke being introduced to the sensing chamber. Is that correct, Mr. Jones?

CEO: Correct.

ATTORNEY: In other words, based on your testimony, it seems clear that the test button is not always a reliable functional test. And those customers who rely solely upon the test button do so at their peril, do they not? And if the Smith family tested your smoke detector in the manner recommended in your user's manual and heard the horn, they assumed -- with fatal consequences -- that the detector was okay. Is that not true?

Product Liability
We will leave our squirming CEO at this point, allowing the reader to arrive at his own conclusions regarding the CEO's answer to the attorney's last group of questions and, of course, the jury's verdict.

The foregoing is not the writer's flight of fancy. Given the many residential fires that occur daily, it is likely that cases involving death and injury resulting from home fires have already been heard in the courts. An attorney knowledgeable about smoke detectors can certainly press his case as vigorously and as effectively as has been outlined above. (Possibly, such an attorney is now reading this article, gathering the insights required to win with substantial damages for his client.) ¹In any case, the liability exposure to smoke detector manufacturers should be obvious.

¹ In fact, the author has received calls from attorneys representing clients who were killed or injured in home fires.

16,000,000 Smoke detectors don't work, according to the federal government.