Temperature measurements obtained by the loggers in vitro were compared to a calibrated liquid-in-glass thermometer, which was defined as the gold standard. Twenty-six temperature loggers (DST micro-T®, Star Oddi, Gardabaer, Iceland) were used. The temperature loggers weighed 3.3 g and had a diameter of 8.3 mm and a length of 25.4 mm. The housing material consisted of implantable, biocompatible ceramic material. The temperature loggers were programmed to measure changes in temperature at 10 s intervals. According to the manufacturer, the temperature loggers are capable of recording temperatures from 5.0°C to +45.0°C, with an accuracy of ±0.2°C and a resolution of 0.032°C. The liquid-in-glass thermometer was validated by the Office of Legal Metrology of the State of Hessen, Germany (Type E 0.1/0/50 DIN 12775; verification certificate 5-3636-09).
The devices were placed into the same water bath (F10 – UC, Julabo, Seelbach, Germany). Every 10, min the temperature of the water bath was increased by 1.0°C covering a range from 30.0 to 45.0°C. The values measured were compared to the value of the gold standard in 1-min intervals. Values measured in the 60 s following each temperature adjustment were discarded from analyses to exclude errors due to acclimatization of temperature changes. To ensure precise matching of paired observations (i.e. measured by loggers and liquid-in-glass thermometer), temperature values were read from the liquid-in-glass thermometer using a stopwatch. Data from the temperature loggers were downloaded (Mercury Application Software®, Star Oddi, Gardabaer, Iceland) and entered with the corresponding liquid-in-glass thermometer data into Excel spreadsheets (MS Office 2003; Microsoft Corporation, Redmond, WA, USA).
The second experiment consisted of five privately owned bitches which were non-spayed generally and gynaecologically healthy. An informed consent was obtained from the owner. The bitches were 7 months to 5 years old and represented following breeds: Boxer, English Setter, Rhodesian Ridgeback, German Shepherd Mix and Terrier Mix. The mean weight was 30.2 kg (range, 22.0–40.0 kg). Vaginal inspection was conducted via speculum (Model 63956, length 15 cm, diameter 2 cm, Storz, Tuttlingen, Germany). Vaginal cytology (Haema–Schnellfärbung, Labor + Technik, Berlin, Germany) was conducted before logger application and after logger removal.
To prevent movement and potential displacement, temperature loggers were attached to a progesterone-free modified Controlled Internal Drug Release (CIDR-blank, Pfizer, New York, NY, USA) device for ewes. A transponder (1.4 × 0.2 cm, BackHome Bio Tec-Transponder, Virbac Corporation, Fort Worth, TX, USA) was placed under the silicon lining of the CIDR-blank. Utilizing a reading device (Minimax II®, Virbac Corporation, Fort Worth, TX, USA), the presence and position of the logger could be confirmed. The temperature loggers were programmed to measure temperatures in 10-min intervals over a 3-day period. The logger was pushed through the speculum using a sterile swab and placed approximately 18 cm deep into the vagina. The dogs were kept in their familiar surroundings with no restrictions to their individual daily exercise routine.
To retrieve the logger, a vaginal endoscopy was performed under carbon dioxide insufflation with a rigid endoscope (S 301-4037, Dr Fritz, Tuttlingen, Germany) consisting of a telescope (4 mm diameter, length 30 cm, 30° angled lens) and a sheath (7 × 5 mm diameter, length 25 cm). The temperature logger attached to the CIDR-blank was pulled out of the vagina by means of flexible biopsy forceps (5 Fr., 40 cm, Storz, Tuttlingen, Germany). Temperature data were downloaded after removal. Rectal temperature measurements were recorded with a digital thermometer (VT 1831®, Microlife USA, Incl, Clearwater, FL, USA) and compared to the vaginal temperature measurements obtained by the logger at the same minute.
Data were analysed using SPSS® for Windows (Version 18.0, IBM Corporation, New York, NY, USA). In the in vitro experiment, the mean difference between the temperature loggers and the liquid-in-glass-thermometer was calculated using a paired t-test. Furthermore, Pearson correlation was determined between the values measured by the loggers and the liquid-in-glass thermometers. The same statistics were performed in the in vivo experiment comparing measures of rectal and vaginal temperature of the bitches. Since using correlation coefficients for clinical measurement comparison can be inappropriate, agreement between the liquid-in-glass thermometer and the logger and vaginal and rectal temperatures, respectively, were analysed graphically for both experiments using the method of Bland and Altman (1986).