Improving Blood Glucose Monitoring for Diabetes
by 

When most people sit down for lunch, the only guidance they seek is from the
menu.  But for an estimated 1 million Americans who are diagnosed with Type I 
diabetes mellitus, lunch--or any other meal--starts with a test using a 
portable device called the blood glucose meter. 

Self-monitoring of blood glucose (a sugar) is called SMBG for short. Most 
insulin-dependent diabetics carry out SMBG four to seven times a day, as it is
crucial to treatment of the potentially degenerative disease of diabetes. 
Properly executed, SMBG enables the users to modify the type of food they eat,
vary the amount and time of exercise, and adjust medications.  SMBG, which is 
also recommended for some diabetics who have non-insulin-dependent diabetes 
(known as Type II), enables the physician and patient to clearly define 
treatment goals and to measure the effectiveness of therapy with minimum
disruption of daily life. 

Yet, some diabetics have had difficulty following the procedures involved with
self-monitoring of blood glucose.  Complaints, which started reaching the Food
and Drug Administration's Center for Devices and Radiological Health in the 
mid-80s, raised questions about the accuracy of blood glucose readings
obtained by diabetics at home, and caused concern about the impact on 
treatment regimens.  For the past three years, FDA has been trying to find out
why errors occur in SMBG--and the search is now almost finished.

                                 Major Advance

Diabetes, which in some forms is genetically linked, reduces the level of 
insulin needed to deliver energy-rich glucose to the cells of the body.  In 
its mildest form, Type II diabetes can go practically unnoticed--an estimated 
5 million American diabetics have never been diagnosed and are presumably 
unaware of their illness or of the long-term health consequences.  But the
most severe type of diabetes mellitus, Type I, characterized by insulin 
dependence, can lead to blindness, gangrene of the extremities, kidney
dysfunction, arteriosclerotic heart disease, and, if these complications are
untreated, death. 

For the 10 percent of America's 11 million diabetics--and a similar proportion
of the 200 million diabetics worldwide--who closely watch their glucose 
levels, blood glucose meters represent a major advance.  Traditionally, a 
urine test strip was used at home by diabetics to monitor glucose levels. 
However, its results were indirect, measuring spillover of sugar in the urine,
and for Type I diabetics it was of only limited use for total diabetes
management.  Generally, only high glucose levels could be detected. 

Blood glucose measurements were introduced for home use in the late 1960s,
enabling diabetics to detect high and low blood sugar levels by visually
noting color changes on a chemical test strip from a single drop of blood.

By the late 1970s, the evolution to a portable meter to "read" the chemical 
strips was complete, and a new trend in diabetes management was born. The 
blood glucose meters, sold over the counter in pharmacies and grocery stores
today, detect the glucose level in the blood on the strip and provide 
immediate warning about the onset of hyper- or hypoglycemia (high or low blood
sugar).  Experts agree that measuring glucose in the blood is a more preferred
method than measuring it in the urine.

Moreover, the meters include many features designed to make self-monitoring 
convenient.  Many operate on batteries and are small enough to fit in a purse 
or a shirt pocket and can be used almost anywhere.  Some contain electronic 
memory, and more advanced models even have built-in modems for transmitting 
the test results to the diabetic's physician. 

                                 The Procedure

With most of the meters, the process of SMBG takes just a few minutes.  For 
some meters, the first step is to calibrate the meter for the particular batch
of test strips being used.  Then, as with the use of visual test strips, the
user obtains a drop of blood by pricking a clean finger with a puncture device
and places that drop of blood on the chemically treated pad of a testing
strip.  The user starts the meter, which measures the exact time necessary for
the chemicals in the pad to react to the blood.  Several seconds later, the 
meter alerts the user to blot or wipe the excess blood from the pad of the
test strip.  At another signal from the meter, the user inserts the strip in
the meter.  Shortly afterward, the display window shows the glucose level in
the blood.  Some older SMBG systems require wiping or blotting of blood from
the test strips, but new technology has eliminated this step from the 
procedure.  For these new meters, once the blood is placed on the test area,
the meter times the reaction and delivers a glucose reading.

                                  Complaints

While SMBG has proved its worth for hundreds of thousands of diabetics, there 
have been many complaints about inaccurate results and difficult meter
maintenance.  The manufacturers of the meters are obligated by law to report
such occurrences to FDA, which regulates medical devices and monitors medical 
devices that present problems for users.

User error seems to be involved in many complaints about inaccurate results.
Various age- and diabetes-linked physical problems of the l million users of
blood glucose meters raise further questions about the number of diabetics who
might have difficulty in performing  the steps for the use of the meter.
According to the American Diabetes Association, about 27 percent of diabetics 
are visually impaired, and 25 percent have loss of sensation in hands or feet 
or problems in moving fingers.  Many diabetics have a low education level with
below average reading skills.  In addition, a substantial number of diabetic
patients are Hispanics with a limited command of English. 

As Joseph S. Arcarese, director of CDRH's Office of Training and Assistance 
noted, "the reports of inaccurate readings with the use of blood glucose
meters are the largest medical device user problem reported to FDA.  We know
something is going wrong.  The question is, what?"

                              Search for Answers

An important early step in the search for answers was a special November 1986 
conference of organizations most directly involved in the problems of 
diabetics--the American Diabetes Association (ADA), FDA, the Centers for
Disease Control, and the National Institute of Diabetes and Digestive and 
Kidney Diseases.  This Consensus Development Conference on Self-Monitoring of 
Blood Glucose, attended by many health professionals and representatives from 
the medical device industry, probed questions about the usefulness and
efficacy of SMBG under conditions ranging from use in hospitals to use in 
summer camps by children with diabetes. 

The principal conclusion, according to John A. Colwell, M.D., Ph.D., the
president of ADA, who chaired the conference, was that SMBG was "recommended
for insulin-treated diabetics [generally, Type I]."  In the view of the health
professionals and regulators of the health industry, the test kits had proved 
their potential therapeutic value, and physicians had encouraged many Type II 
diabetics to use SMBG to monitor their blood glucose. 

The second most important finding of the conference was that the blood glucose
meters should be made easier to use, and that the users should be better
trained in their operation.  "We did not have much [research] data," Dr.
Colwell recalled, "but we felt that there was a tremendous need for better
user education."

                                  FDA Studies 

Following the consensus conference, FDA conducted several in-house studies. 
One probed the reliability, accuracy and precision of 14 marketed portable
meters.  An analysis of the meters by FDA's laboratory in Baltimore showed
that, with minor exceptions, they performed as they were supposed to when used
according to the manufacturer's directions.  But an in-house FDA study of 
instructions included by the manufacturers with the meters strongly suggested 
that much of the material did not adequately give the users the 
self-monitoring guidance they needed. 

The specific problems faced by diabetics emerged only from a detailed user
analysis of a sample of nine meters and their accompanying instructional
materials by the Pacific Science and Engineering Group in an FDA-sponsored
study, "Human Factors Analysis of Blood Glucose Monitoring."  With the final
report near completion, preliminary conclusions show that FDA was on the right
track.  In the FDA findings, approximately two-thirds of the experienced users
made significant errors.  The main factors behind the inaccurate readings 
appear to be inadequate training in the use of meters and misunderstanding of 
the manufacturers' instructions leading to errors in the operation of the 
meter.

Common user errors were:

* failure to follow the manufacturer's procedures 
* inadequate amount and placement of blood on the test strip
* failure to calibrate the meter
* lack of cleaning
* use of outdated strips. 

                                 Meter Design 

The devices in the FDA-sponsored study included the widely used meters
equipped with reflectance photometers, which measure light reflected from a 
reagent strip after it has undergone a chemical reaction.  Microcurrent 
meters, which use a new technology to gauge the electric current produced by
the blood resting on the tip of the test strip, were also included in the 
study.

Liquid crystal displays, which are crucial to the successful use of the 
meters, were found almost uniformly to perform within acceptable limits.  All 
displays were found to be legible even when viewed in direct sunlight.

But most of the meters contained engineering and design characteristics that
made SMBG complicated for the users.  Some control or button labels were too
difficult to read, and some of the buttons were not arranged in the order in
which they are routinely used.

More problematic were the relatively faint beeps that alert the user to the 
various steps in the monitoring process.  Diabetics with poor vision may have 
difficulty reading the instructions on the display and thus rely more heavily 
on the audible tones of the meter.  According to the survey, "tones produced
by one half of the meters could not be readily heard over a normal
conversation five feet away."  Another problem identified by the human factors
study was that routine replacement of batteries required "too much fine motor 
coordination and was quite difficult for some users." 

                              Instruction Manuals 

The 11 manuals analyzed by FDA presented still more obstacles for at least
some users of the meters.  Some were printed in letters smaller than the 10 to
12 point size  recommended for the greatest reading ease.  (For comparison, 
the letter size on this page is 10 points.)  Graphics--especially important 
for people who have poor reading skills--were relatively scant.  FDA staff
believe that for most of the manuals, improved nonverbal instructions would be
helpful.

The scarcity of explanatory pictures was particularly troubling because of the
overall low readability scores of the manuals.  As the study notes, 
information in instructions should be "inherently clear to the user," and 
medical or technical terms should be minimized.  Yet, phrases such as 
"programming sets the instrument's electronics to match the reactivity" or
"all informational digits are functioning" were common in several manuals.

Because the reading comprehension of the average American adult is below the
ninth grade level, it is usually recommended that instructional material be 
written at the sixth-grade level.  The reading-grade level of the examined
manuals ranged from the late seventh grade, with an average score of late 
ninth grade.

"Efforts are needed to improve the readability of most manuals," the
preliminary study concluded.  Another important recommendation called for the 
inclusion of information about the most common testing pitfalls and how to
correct them. 

                             Corrections Under Way

After a three-year search by FDA to identify the causes of user errors in 
SMBG, remedies are under way. Several manufacturers of blood glucose meters 
have already begun improving features of the meters and instructional 
materials identified by the San Diego researchers.  Complicated procedures are
being reduced, and instruction manuals, tested by meter users, have improved
formats with clear illustrations.  Manufacturers' 800 phone numbers are now 
available to help users with problems in SMBG.

To zero in on the users' mistakes in the monitoring, FDA plans a workshop in
May 1990 with health professionals, meter manufacturers, and other groups 
involved in the care of diabetics.  The goal of the workshop is to develop
SMBG educational and training strategies that would be made available 
nationwide to diabetics.

--Contributors to this article were Susan Meadows, project officer for the
SMBG study, and Mike Kubic of FDA's public affairs staff. 

                             Meter Monitoring Tips

If you suspect a reading is in error: 

* Check the calibration of the meter. 
* Conduct a control test. 
* Repeat your blood glucose test. 
* Call a diabetes health professional or the meter manufacturer's 800 number
for help. 

Tips for reducing user errors:

* Obtain professional training and guidance.
* Use fresh strips and supplies.
* Clean your meter frequently.
* Check your SMBG technique routinely with a professional.
* Follow the manufacturer's instructions carefully.