Although the mercury sphygmomanometer has been considered the gold standard for the indirect measurement of blood pressure (BP),1 in recent years, some public health advocates have raised concerns about the use of mercury in medical practice.2 These concerns have prompted many medical facilities to replace mercury manometers with aneroid devices. Critics of nonmercury sphygmomanometers assert that insufficient data exist to validate the use of aneroid sphygmomanometers in a diverse range of patients, BP levels, and clinical settings over an extended period of time,1 but studies performed in university hospitals and clinics have documented the accuracy of nonmercury devices. The present report examined the performance of these devices in outpatient medical practices.
The mercury sphygmomanometer is the undisputed gold standard for the indirect measurement of blood pressure. Some public health advocates have recently expressed concern about the use of mercury in medical practice.2 This concern has prompted many medical facilities to replace mercury manometers with aneroid devices. The present report examined the performance of 282 aneroid sphygmomanometers in outpatient medical practices. Results were examined for predetermined end points within ±3 mm Hg from the reference values and to indicate zero at no pressure. Ninety-three devices (33%) failed to perform at ≥1 pressure levels. Most (76%) of the failures were due to low readings. Only 7 of the 93 failing units did not rest at zero, making this an unreliable indicator of accuracy. Inaccurate readings of aneroid sphygmomanometers may result in a failure to diagnose and treat hypertension, thereby placing hypertensive patients at risk for end-organ damage and cardiovascular events.
Medical practices were contacted by phone and permission was obtained to evaluate aneroid sphygmomanometers used in their offices. These practices represented a variety of clinical specialties including general medicine, pediatrics, cardiology, nephrology, obstetrics/gynecology, and surgery. The practices included a hospital clinic, community hospital clinics, a community health center, and several private practices. Two trained investigators compared BP measurements determined with an aneroid instrument and a standard mercury manometer (Baumanometer; W. A. Baum Co, Inc, Copiague, NY). These measurements were performed with a Y-shaped tube. The inflation bulb was attached to the base of the Y and the 2 branching arms connected to the mercury and aneroid devices. One investigator randomly adjusted the pressure in the mercury manometer to 50, 100, 150, 200, or 250 mm Hg while the second blinded investigator recorded the BP reading from the aneroid instrument. Each device was also examined with no pressure applied to see whether its needle rested either at 0±1 mm Hg or within a “zero box” on the face of the dial.
Consistent with the American National Standards Institute (ANSI) and the Association for the Advancement of Medical Instrumentation,3 an aneroid device was considered accurate at a designated pressure if it agreed with the mercury device within 3 mm Hg. Results were examined for predetermined end points: the overall failure rate (defined as >±3 mm Hg from the reference value at ≥1 pressure levels), failure rates at specific pressure levels, mean ± SD from the reference value at each pressure interval, failure to indicate 0 at no pressure, failure rates with portable vs wall-mounted units, failure rates by manufacturer, and failure rates by physician specialty. No data were available on how long the individual units had been in use. Statistical significance was tested by chi-square test.
We examined 282 aneroid sphygmomanometers. Ninety-three (33.0%) failed to perform within the ±3 mm Hg standard at ≥1 pressure levels. Sixty devices (21.3%) failed at >1 pressure level, including 40 (14.2%) that failed at all 5 levels for a total of 286 failed readings in 93 devices. Only 79 of 286 failures (27.6%) were due to high readings, whereas the majority (76.0%) were due to undermeasurement of BP. The highest failure rates tended to occur at the following pressures: 50 mm Hg (22.0%), 100 mm Hg (19.0%), and 150 mm Hg (19.0%).
Thirty-five units (12.4%) failed to rest either at 0±1 mm Hg or within the “zero box” with no pressure. Only 7 of the 93 failing units discussed above did not rest at zero. This was not a reliable indicator of an aneroid unit that would be inaccurate at inflated pressures; however, all 7 of the instruments that failed at all pressure levels also failed to rest at zero.
One hundred sixteen devices were portable, and 166 were wall-mounted. Forty-six of the failing units were portable, corresponding with a failure rate of 39.7%. These included 6 instruments that did not rest at zero. Wall-mounted units accounted for 57 of the failures, corresponding with a failure rate of 34.3%. This difference between portable and wall-mounted units was statistically significant (39.7% vs 34.3%; χ2=0.867; P=.05). One wall-mounted unit (0.6%) failed to rest at zero vs 6 (5.2%) portable units.
Most of the aneroid devices examined were of 2 brands: Trimline (Trimline Medical Products Corp, Branchburg, NJ) or Tycos (Welch Allyn, Inc, Skaneateles Falls, NY). The Trimline device demonstrated a failure rate of 29.5% (36/122 units). The Tycos device demonstrated a failure rate of 42.0%(50/119 units). This difference was statistically significant (29.5 vs 42.0%; χ2=4.106; P=.05). There were other brand names that were not included in this subgroup analysis. The difference between brands was not caused by the type of device, because the percent failure for the Tycos portable instruments was 36.0% vs 30.0% for Trimline (χ2=0.294; P=.05) and the failure rate for wall-mounted instruments was 44% for Tycos and 28% for Trimline (χ2=4.153; P=.05).
The highest failure rate occurred in private medical offices (47/101 or 46.5%). A hospital-based gynecology clinic had a failure rate of 31.8% (7/22). The failure rates in hospital-based internal medicine, medical subspecialty, and nonmedical specialty clinics were 18.0%, 24.0%, and 31.0%, respectively.
The present results suggest that 33.0% of aneroid devices used to measure BP in outpatient clinical practices do not meet performance standards required for these devices. The average magnitude of deviation from the mercury derived value and the absolute deviation for most devices was usually small. However, at 250 mm Hg, aneroid readings ranged from 240 to 260; at 200, from 186 to 210; at 150, 142 to 198; at 100, 86 to 107, and at 50, 43 to 57 (Figure).
Aneroid instruments present a significant calibration challenge because even minor trauma can adversely affect their calibration. Consequently, most aneroid manufacturers recommend semiannual calibration with a mercury manometer, but few institutions, and even fewer physician's offices, have implemented programs to evaluate and calibrate these instruments. Furthermore, an additional disincentive is the fact that most aneroid devices must be returned to the manufacturer for calibration.4 Mercury sphygmomanometers, on the other hand, are durable, infrequently require repair, and have been validated against direct intra-arterial pressure measurements.5
Two recent studies have validated the use of aneroid instruments. Canzanello and associates6 reported that “virtually 100%” of values obtained from aneroid devices at the Mayo Clinic were within 4 mm Hg of their reference value. Most of these values were obtained from annually inspected inpatient devices, however. Only 8.6% of the devices (3 out of 35)6 examined were less rigorously maintained than the devices in the Mayo outpatient clinics, and 1.4% of these devices (4 out of 283) required replacement because of readings >4 mm Hg from the reference value.6 Furthermore, ±4 mm Hg exceeds the ±3 mm Hg recommended for sphygmomanometers by the ANSI and the AAMI. Yarows and Qian7 reported a 4.4% aneroid device failure rate at the University of Michigan using ±3 mm Hg as the standard. The failure rate rose to 16.2% using ±2 mm Hg as the standard for device failure, however, as advocated by Yarows and Qian as well as the British Hypertension Society.7,8 Earlier studies of aneroid sphygmomanometers have shown even higher failure rates, typically 30% to 40% (Table).9–11 Our results demonstrate an overall failure rate of 33.0% greater than that of the Mayo Clinic or University of Michigan studies but less than those of earlier studies. The improvement in device performance since early studies of aneroid devices may be due to changes in manufacturing techniques and increased awareness of the need for regular calibration of the aneroid instruments.
|Study||Failure Standard||Failure Rate|
|Canzanello et al, 20016||>4 mm Hg or failure to rest at zero||1.4%|
|Yarows and Qian, 20017||>3 mm Hg||4.4%|
|>2 mm Hg||16.2%|
|Mion and Pierin, 199812||>3 mm Hg||58.0%|
|Bailey et al, 199111||>3 mm Hg at ≥2 pressures||34.8%|
|McKay et al, 199010||>4 mm Hg||40.0%|
|Burke et al, 19829||>4 mm Hg||30.0%|
Comparison among studies is complicated by the definition of instrument failure. The ANSI and the AAMI advocate a standard for aneroid sphygmomanometers of ±3 mm Hg, the definition used in the present study. Clinical research results support a standard at least this rigorous. The British Hypertension Society suggests that ±2 mm Hg is a more appropriate standard for judging the performance of sphygmomanometers. Applying the ±2 mm Hg standard to the present report would increase the failure rate of aneroid devices to 85.0%.
Most (76.0%) of failures in the present study were due to underreading the pressure. This finding is consistent with other studies suggesting that aneroid instruments underestimate BP.11,12 Low readings may result in failure to diagnosis and treat hypertension adequately, thereby placing hypertensive patients at increased risk for end-organ damage and cardiovascular events.
Whether an instrument rested at zero with no pressure applied showed little correlation with its overall performance. All instruments that were inaccurate at all 5 pressure levels also failed to rest at zero, but almost half of the inaccurate instruments did rest at zero. Thus, while the failure of an instrument to rest at zero should alert its user to its need for calibration, resting at zero is not predictive of accuracy.
Tycos instruments, despite their inaccuracy, were more likely to rest at zero. This difference underscores the fact that resting at zero is not predictive of accuracy. The tendency of the Tycos instrument to rest at zero more often is almost certainly related to its dial design. Tycos dials possess a relatively wide “zero box” on the face. If the needle rests within the box when no pressure is applied, the unit is considered to rest at zero. Trimline dials have only a single line representing zero.
The performance of sphygmomanometers varied by the type of medical practice. Private offices demonstrated a failure rate (46.0%) 3 times that of hospital-based practices (18.0%). This disparity between hospital- and community-based practices may reflect greater awareness and personnel in hospitals to address the issue of calibration. Portable instruments are also subjected to more movement and trauma. In our analysis, we did not take into consideration the age of the device. It is one of the limitations of this study, since it would seem that older devices (with more use and abuse) in private settings would be more likely to be inaccurate. We do not have comparative data on how long the Tycos and Trimline units had been in use, and this variable could have affected the difference in their accuracy.
Our results suggest that portable devices are more vulnerable to inaccurate BP measurement and may require more frequent calibration than wall-mounted devices. Detection of measurement inaccuracy required actual calibration of the device and could not be identified by inspection of the instrument and its zero indication alone. There is presently a movement to replace mercury devices in medical practice with aneroid instruments. This must be accompanied by programs to ensure periodic calibration of these replacement instruments, especially in office practices, where the capacity for this activity may be less than in the hospital clinic setting.
The findings of this study illustrate important challenges related to BP measurement. Replacing mercury sphygmomanometers with aneroid devices involves the substitution of a superior instrument with a less reliable alternative. Aneroid sphygmomanometers require vigilance if their proper calibration is to be maintained. A periodic program of evaluation and calibration should be instituted in all medical practices utilizing them. Simply noting whether a device rests at zero is inadequate. Certain brands or portable instruments may require more frequent evaluation. The profound consequences of inadequate diagnosis or treatment of hypertension require these rigorous measures.