Anemia affects 40% of pregnant women worldwide [2
], and is detrimental to the health of mother and their fetus. Hemoglobin concentration is used for detecting anemia and also evaluating the gravity of this disorder. The use of a laboratory hematology analyzer is problematic in remote areas. Portable hemoglobin meters offer a rapid, handy and inexpensive measurement of hemoglobin. In addition, the finger puncture to allow capillary blood sampling as an easy technique is less resource-intensive than vein puncture, and is more acceptable to patients and the community. The present study evaluated the agreement between the HemoCue device and a laboratory hematology analyzer for capillary hemoglobin measurement. This is the first validation study conducted in a high altitude area, thus filling a gap in this area.
The 95% LOA in present study is much wider compared to a study (-2.93 g/l and 3.38 g/l) carried out in infants in Honduras [6
] and a study (-3.7 g/l and 4.5 g/l) carried out in patients undergoing aortic surgery [18
]. However, several studies have demonstrated a similar LOA to that obtained in this study. Bhaskaram et al. demonstrated a LOA of -33 g/l and 9 g/l in India [19
]. Rippmann et al. found a LOA of -6.0 g/l and 18.0 g/l in surgical blood samples [8
], while a LOA of -18.0 g/l and 16.8 g/l in a study by van de Louw et al. was obtained in patients gastrointestinal bleeding in France [20
In our study 21.7% of the absolute values of the difference were above 10.0 g/l. These results are slightly poorer than those reported in previous studies. Radtke et al. reported a difference of greater than 10.0 g/l in 9% of samples from unselected German blood donors [21
], and in a study by Rippmann et al. [8
], and the figure is 17.9%. van de Louw et al. found 21% of the difference was greater than 10.0 g/l [20
]. The results in our study and from the previous studies highlight the need for a pilot validity study before the HemoCue device can be used in a particular population or setting such as high altitude regions. The determination of hemoglobin in capillary blood using the HemoCue device was not accurate in our study, which was similar to validation studies in patients and blood donors. The result consists with a study conducted in Mexico middle high altitude area [11
]. We therefore do not recommend this device to measure capillary hemoglobin in pregnancy in high altitude regions due to the discrepancies associated with this device compared to the laboratory hematology analyzer.
ICC and CCC are commonly used to evaluate agreement between measurements. In our study the ICC was 0.795, and the CCC was 0.793. These findings were similar or slightly poorer than those reported in previous studies in general settings [6
]. These findings together with the LOA, confirm that capillary hemoglobin determination using the HemoCue does not have an acceptable agreement with capillary hemoglobin determination using the laboratory hematology analyzer.
In a study in Indonesian mothers, Sari et al. reported a sensitivity of 70.6% and a specificity of 97.5% in detecting anemia [23
]. Radtke et al. reported a sensitivity of 98.0% and a specificity of 50.0% in screening blood donors [21
], and Sawant et al. found a sensitivity of 99.0% and a specificity of 45.0% in screening blood donors [22
]. In another study in Mexico, Neufeld et al. reported a sensitivity of 79% in detecting anemia in adults and 84% in children, and a specificity of 97% and 93% [11
], respectively. In the present study, setting the laboratory analyzer as reference method, the sensitivity and specificity for anemia screening were 94.9% and 76.7%, the positive predictive value was 84.1% and the negative predictive value was 92.0%. These results are similar to or slightly better than previous studies, and the results seem acceptable.
Several authors have pointed out that capillary blood samples are more variable than venous blood samples [6
], which may partly explain the poor agreement observed in our study. A shortcoming of this study is that this study did not measure with both methods using the same puncture. Disagreement may also be affected by the quantity of capillary blood samples, although the only one nurse who conducted the sampling and measurement process had undergone strict trainings.
The sample size in our study meets the general recommendations of Altman of at least 50 subjects in a methods comparison study [24
This study is the first to evaluate HemoCue device in the determination of capillary blood hemoglobin in pregnant women in a very high altitude area. We also obtained detailed information on the agreement between the portable HemoCue and a laboratory analyzer on hemoglobin levels in pregnancy. Due to poor acceptance among participants, we were unable to obtain duplicate measurements for both methods to assess the reliability of these methods. However, the laboratory analyzer has been reported to perform very well in other published articles. Although we obtained poor agreement between HemoCue device and the laboratory analyzer for the determination of capillary blood hemoglobin, we do not know whether the high altitude in the study area contributed to this poor agreement. This is a topic that needs investigation.