TY - JOUR
AU - Garrett,, Tim
AB - Abstract Purpose The temperature profiles of antibiotic-containing elastomeric infusion devices used by ambulatory care patients under various environmental conditions were evaluated. Methods A prospective, descriptive survey of temperature exposure was conducted in 4 publically funded hospitals. Over a 12-month period, electronic temperature-recording devices were attached to the antibiotic infusion devices (infusers) of prospectively randomized hospital-in-the-home (HITH) participants. Temperatures were recorded immediately after infuser connection and every 5 minutes thereafter for 24 hours. A structured data collection form was used to collect information on basic clinical and demographic characteristics and aspects of daily living (i.e., how and where the infuser was carried during the day, times the participant went to and arose from bed, location of the infuser while sleeping, and dates and times the infuser was connected and disconnected). Results A total of 115 patients successfully completed the study (17–91 years old, 55% males). A total of 31,298 temperature readings were collected. The storage location of the infuser did not influence daytime readings. However, the overnight storage location did have a significant impact on the temperatures recorded overnight. The mean temperatures of infusers stored on the bed or on the body overnight were significantly higher than those for infusers stored away from the bed. Diurnal and seasonal influences were also detected. Significantly warmer temperatures were recorded in afternoons and evenings and during the summer months. Conclusion Antibiotics administered to HITH patients via continuous infusion were frequently exposed to temperatures in excess of 25 °C. Specific patient behaviors and seasonal and chronological factors influenced temperatures. The findings challenge the validity of current fixed-temperature models for testing stability, which do not reflect conditions found in clinical practice. ambulatory, antibiotic, elastomeric, infusion, stability, temperature Key Points This is the first study that has measured the real-time temperature of i.v. infusion devices in hospital-in-the-home (HITH) patients, spanning all seasons over a 12-month period. Controlling the location of infusers during the night and other patient behaviors can significantly reduce mean daily temperatures and reduce out-of-range temperature spikes, potentially improving antibiotic stability. These findings challenge the appropriateness of current fixed-temperature stability testing, which is not representative of clinical practice in outpatient-based HITH patients. There has been significant growth in the development of hospital-in-the-home (HITH) and outpatient parenteral antimicrobial therapy (OPAT) programs that allow the administration of i.v. antimicrobials outside of the hospital setting. These programs are used to treat patients with community-acquired infections, hospital-acquired infections, and other serious, recurrent, or multidrug-resistant infections. In most cases, HITH or OPAT programs utilize i.v. delivery devices to administer antimicrobial therapy, either intermittently or continuously, to patients over 24 hours and therefore allow for once-daily attendance of nursing staff to connect the antibiotic solution and perform clinical checks. This practice raises the issue of the stability of all i.v. medications, particularly i.v. antibiotics, over 24 hours. Some HITH programs use twice-daily visits to overcome some of the stability concerns, but this practice affects program efficiency. It is well recognized that the stability of an antibiotic is influenced by temperature, drug concentration, and pH. Generally, an antibiotic is considered suitable for i.v. administration if less than 10% of the antibiotic in solution has degraded by the end of the infusion period (usually 24 hours).1,2 Most i.v. antibiotics in solution display concentration- and temperature-dependent degradation, and these factors should be taken into consideration when assessing an antibiotic’s suitability for home-based administration. Currently, the suitability of an antibiotic in solution is usually determined from in vitro studies using beyond-use dates based on fixed-temperature stability testing models, in which degradation rates are measured at a continuous static room temperature, defined by Trissel1 as 21–25 °C. However, the ambient temperatures to which patients are exposed in community settings have not been extensively reviewed. It could be argued that greater temperature fluctuations could be experienced outside the controlled environment of the hospital or laboratory and that these temperature fluctuations may affect antibiotic integrity. Some small-scale studies found that temperatures of up to 37 °C have been recorded on portable infusion devices while in use by patients.3,4 This observation has led to multiple antibiotic stability studies using higher fixed-rate temperatures, ranging from 25 to 37 °C.5–12 However, no large study has investigated the operating temperatures of portable i.v. infusion devices when used in nonhospital environments. Understanding the real-life fluctuations in i.v. infusion device temperatures and antibiotic integrity may be important when considering the suitability of an i.v. antibiotic for home-based administration. This study was conducted to determine the temperature profile of antibiotic infusions while in use by a group of HITH patients over a 24-hour period. Factors that contribute to any observed variations in temperatures were also identified. Approval for the study was obtained from the Northern Sydney Central Coast Human Research and Ethics Committee. Methods Study setting. The Central Coast region is the third largest urban area within the Australian state of New South Wales and has a population of close to 320,000.13 Health services are provided via the Central Coast Local Health District (CCLHD), encompassing principal referral hospitals, major metropolitan hospitals, and subacute care hospitals. The CCLHD provided a total of 78,419 episodes of care from July 2012 through June 2013. HITH services are provided by the CCLHD Acute Post-Acute Care (APAC) team. The provision of OPAT via elastomeric infusion devices is a significant component of the APAC service. Study design and population. Patients served by the CCLHD who received home-based antibiotic infusions administered via a portable elastomeric i.v. infusion device between July 1, 2012, and June 30, 2013, were eligible for study inclusion. Patients were provided with written and oral information about the study before enrollment, and informed consent was obtained from all participants. There were no differences in the type of therapeutic intervention received or any other aspect of treatment between participants and nonparticipants. A maximum of 4 eligible respondents were randomly selected each week for study inclusion using an online randomization tool.14 Participants with multiple admissions were eligible to participate only once in the study. The elastomeric i.v. infusion devices, or infusers, (Infusor, Baxter Healthcare) used were nonelectronic medication pumps designed to provide ambulatory continuous i.v. infusion therapy. Four computerized temperature-recording devices or data loggers (Tinytag Transit 2 Gemini data loggers, Chichester, United Kingdom), capable of recording temperatures of −40 to 70 °C with a variance in accuracy of ±0.4 °C, were configured to record ambient temperatures at continuous 5-minute intervals. A data logger was attached to each participant’s infuser using surgical tape before infuser distribution (Figure 1). Figure 1 Open in new tabDownload slide Elastomeric infusion device with temperature recorder device (yellow object with attached numbered tag) secured with surgical tape. Figure 1 Open in new tabDownload slide Elastomeric infusion device with temperature recorder device (yellow object with attached numbered tag) secured with surgical tape. A structured data collection form used open-ended questions to collect information on basic clinical and demographic characteristics and aspects of daily living (i.e., how and where the infuser was carried during the day, times the participant went to and arose from bed, location of the infuser while sleeping, and dates and times the infuser was connected and disconnected). The nurse visited participants at their homes on two occasions approximately 24 hours apart to connect and disconnect infusers; conduct routine clinical observations; distribute, check, and collect the data collection forms; and retrieve the data logger. Temperature readings were measured and recorded at 5-minute intervals for the 24 hours after connection of the infuser. Statistical analysis. Data were analyzed using Stata software, version 13 (StataCorp LLC, College Station, TX). Group differences were assessed using chi-square analyses and Kruskal-Wallis tests. Post hoc comparisons of group medians were conducted using Mann-Whitney U tests, with a Bonferroni correction for multiple comparisons. The effects of time were examined using generalized linear mixed models (GLMMs). The a priori level of significance was 0.05. Results. Of the 161 total patient episodes that occurred during the study period, 115 patients participated in the study. The mean ± S.D. age of participants was 58 ± 17 years (range, 17–91 years); 63 participants (55%) were male. Table 1 presents the clinical indications for antibiotic therapy, the antibiotic agents used, the locations the infuser was carried during the day and stored at night, and the seasons when treatment was conducted. Bone and joint infections and skin and soft tissue infections were the predominant indications for antibiotic therapy. Flucloxacillin, vancomycin, piperacillin–tazobactam, and cefazolin were the antibiotics most commonly administered. Table 1 Therapy and Study Characteristics (n = 115) Characteristic No. (%) Indication for antibiotic  Bacteremia or sepsis 6 (5.2)  Bone or joint infection 70 (60.9)  Endocarditis 3 (2.6)  Respiratory infection 9 (7.8)  Skin or soft tissue Infection 23 (20.0)  Other 4 (3.5) Antibiotic used  Benzylpenicillin (buffered) 9 (7.8)  Cefazolin 14 (12.2)  Ceftazidime 3 (2.6)  Clindamycin 1 (0.9)  Flucloxacillin 38 (33.0)  Piperacillin–tazobactam 19 (16.5)  Ticarcillin–clavulanic acid 6 (5.2)  Vancomycin 25 (21.7) Infusor location during daytime use  Carried on body 86 (74.8)  Hung on body 9 (7.8)  Carried off body 8 (7.0)  Hung off body 5 (4.3)  Not specified 7 (6.1) Infusor location when sleeping  On body or in bed 24 (20.9)  On bed 56 (48.7)  Away from bed 27 (23.5)  Not specified 8 (7.0) Season when study performed  Spring 25 (21.7)  Summer 18 (15.7)  Autumn 31 (27.0)  Winter 41 (35.7) Characteristic No. (%) Indication for antibiotic  Bacteremia or sepsis 6 (5.2)  Bone or joint infection 70 (60.9)  Endocarditis 3 (2.6)  Respiratory infection 9 (7.8)  Skin or soft tissue Infection 23 (20.0)  Other 4 (3.5) Antibiotic used  Benzylpenicillin (buffered) 9 (7.8)  Cefazolin 14 (12.2)  Ceftazidime 3 (2.6)  Clindamycin 1 (0.9)  Flucloxacillin 38 (33.0)  Piperacillin–tazobactam 19 (16.5)  Ticarcillin–clavulanic acid 6 (5.2)  Vancomycin 25 (21.7) Infusor location during daytime use  Carried on body 86 (74.8)  Hung on body 9 (7.8)  Carried off body 8 (7.0)  Hung off body 5 (4.3)  Not specified 7 (6.1) Infusor location when sleeping  On body or in bed 24 (20.9)  On bed 56 (48.7)  Away from bed 27 (23.5)  Not specified 8 (7.0) Season when study performed  Spring 25 (21.7)  Summer 18 (15.7)  Autumn 31 (27.0)  Winter 41 (35.7) Open in new tab Table 1 Therapy and Study Characteristics (n = 115) Characteristic No. (%) Indication for antibiotic  Bacteremia or sepsis 6 (5.2)  Bone or joint infection 70 (60.9)  Endocarditis 3 (2.6)  Respiratory infection 9 (7.8)  Skin or soft tissue Infection 23 (20.0)  Other 4 (3.5) Antibiotic used  Benzylpenicillin (buffered) 9 (7.8)  Cefazolin 14 (12.2)  Ceftazidime 3 (2.6)  Clindamycin 1 (0.9)  Flucloxacillin 38 (33.0)  Piperacillin–tazobactam 19 (16.5)  Ticarcillin–clavulanic acid 6 (5.2)  Vancomycin 25 (21.7) Infusor location during daytime use  Carried on body 86 (74.8)  Hung on body 9 (7.8)  Carried off body 8 (7.0)  Hung off body 5 (4.3)  Not specified 7 (6.1) Infusor location when sleeping  On body or in bed 24 (20.9)  On bed 56 (48.7)  Away from bed 27 (23.5)  Not specified 8 (7.0) Season when study performed  Spring 25 (21.7)  Summer 18 (15.7)  Autumn 31 (27.0)  Winter 41 (35.7) Characteristic No. (%) Indication for antibiotic  Bacteremia or sepsis 6 (5.2)  Bone or joint infection 70 (60.9)  Endocarditis 3 (2.6)  Respiratory infection 9 (7.8)  Skin or soft tissue Infection 23 (20.0)  Other 4 (3.5) Antibiotic used  Benzylpenicillin (buffered) 9 (7.8)  Cefazolin 14 (12.2)  Ceftazidime 3 (2.6)  Clindamycin 1 (0.9)  Flucloxacillin 38 (33.0)  Piperacillin–tazobactam 19 (16.5)  Ticarcillin–clavulanic acid 6 (5.2)  Vancomycin 25 (21.7) Infusor location during daytime use  Carried on body 86 (74.8)  Hung on body 9 (7.8)  Carried off body 8 (7.0)  Hung off body 5 (4.3)  Not specified 7 (6.1) Infusor location when sleeping  On body or in bed 24 (20.9)  On bed 56 (48.7)  Away from bed 27 (23.5)  Not specified 8 (7.0) Season when study performed  Spring 25 (21.7)  Summer 18 (15.7)  Autumn 31 (27.0)  Winter 41 (35.7) Open in new tab Daytime infuser location. During daytime infuser use, the majority of participants carried the antibiotic infuser “on body” (i.e., in a portable waist pouch situated on the main trunk of the body). A smaller number of participants carried the infuser “hung on body” (i.e., in a portable waist pouch suspended around the neck or over the shoulder) or “hung off body” (i.e., in a portable waist pouch suspended externally to the body, as from a wheelchair or walker). Some patients used the infuser “carried off body” (i.e., without the waist pouch and carried on the periphery of the body, as in a pocket or handbag), and the remainder did not specify where the infuser was carried. Bedtime infuser location. Almost half of all participants reported storing the infuser “on bed” (i.e., on top of bedcovers or above the pillow) while sleeping, with the remainder storing the infuser “away from bed” (i.e., on a bedside table or chair or hung from the bed frame or headboard) or “on body or in bed” (i.e., under the bedcovers with or without the waist pouch) or not specifying a bedtime storage location. Seasons. Winter (June–August) was the most common season for participation in the study, followed by autumn (March–May), spring (September–November), and summer (December–February). Temperature summary. A total of 31,298 temperature readings were collected. The frequency distribution of the readings is presented in Figure 2. Temperatures were normally distributed across the range of 9.3–40.5 °C. The mean ± S.D. temperature was 25.7 ± 4.6 °C, and the median temperature was 26.2 °C (interquartile range, 22.8–29.0 °C). Only 39.8% of readings were 25 °C or below. The remaining 60.2% of readings were greater than 25 °C, with 17.4% and 0.5% higher than 30 and 35 °C, respectively. Figure 2 Open in new tabDownload slide Frequency distribution of all 31,298 temperature readings from all study patients recorded in the 24 hours after antibiotic infuser connection. Density (vertical axis) refers to the number of readings at each temperature (horizontal axis) expressed as a proportion of all temperature readings. Figure 2 Open in new tabDownload slide Frequency distribution of all 31,298 temperature readings from all study patients recorded in the 24 hours after antibiotic infuser connection. Density (vertical axis) refers to the number of readings at each temperature (horizontal axis) expressed as a proportion of all temperature readings. Changes in infuser temperatures over time. The mean temperature readings at the time of infuser connection and at 5-minute intervals for 24 hours after connection are presented in Figure 3. GLMM analyses indicated that mean temperatures changed significantly over time (p < 0.001). After infuser connection, mean temperatures increased steadily from 22.9 °C to a peak of 27.3 °C at 4 hours and 45 minutes. Mean temperatures were then maintained until 8 hours after infuser connection before declining to approximately 25 °C at 14 hours and remaining relatively stable until 22 hours after connection. Figure 3 Open in new tabDownload slide Mean temperature readings at the time the antibiotic infuser device was connected (time 0) and at 5-minute intervals for 24 hours after connection. Figure 3 Open in new tabDownload slide Mean temperature readings at the time the antibiotic infuser device was connected (time 0) and at 5-minute intervals for 24 hours after connection. The median percentages (25th–75th percentile) of readings greater than 25, 30, and 35 °C were 61.2% (32.0–91.6%), 6.9% (0.7–29.8%), and 0.0% (0.0–0.0%), respectively. The temperature profiles of 11 participants included readings that exceeded 35 °C. For these participants, between 2 (0.7%) and 86 (32.6%) readings exceeded 35 °C. Diurnal effects.Figure 4 presents the mean temperature readings 24 hours after infuser connection in real time. A significant effect of time on temperature was detected (GLMM analysis, p < 0.001). Temperature readings during morning (0731–1200) and overnight (2201–3130) periods tended to be lower than afternoon (1201–1630) and evening (1631–2200) periods. Figure 4 Open in new tabDownload slide Mean temperature readings after infuser connection in real time. The span of 0731–1200 was designated as morning, 1201–1630 as afternoon, 1631–2200 as evening, and 2201–3130 as overnight. Times greater than 2400 represent day 2 of the collection period. Figure 4 Open in new tabDownload slide Mean temperature readings after infuser connection in real time. The span of 0731–1200 was designated as morning, 1201–1630 as afternoon, 1631–2200 as evening, and 2201–3130 as overnight. Times greater than 2400 represent day 2 of the collection period. Seasonal effects. Participants were stratified into 4 groups by the season during which they received antibiotic treatment. For each group, the mean temperature readings at 0 minutes (immediately after infuser connection) and at 5-minute intervals after connection are presented in Figure 5. A significant difference in mean temperatures was noted across groups (GLMM analysis, p < 0.001). The mean temperatures recorded in summer were higher than those recorded in spring, autumn, or winter. Figure 5 Open in new tabDownload slide Mean infuser temperature at 5-minute intervals for 24 hours after connection of antibiotic infuser (time 0). Data are grouped on the basis of whether antibiotic treatment was received in spring, summer, autumn, or winter. Figure 5 Open in new tabDownload slide Mean infuser temperature at 5-minute intervals for 24 hours after connection of antibiotic infuser (time 0). Data are grouped on the basis of whether antibiotic treatment was received in spring, summer, autumn, or winter. Post hoc comparisons revealed a significant difference (p < 0.001) across seasons in the median percentage of readings of >25 and >30 °C (Table 2). No significant difference was found in the median percentage of readings of >35 °C among seasons (p = 0.858). Table 2 Percentage of Infuser Temperature Readings Exceeding 25 and 30 °C Across the Seasons Season Median % Readings Higher Than Indicated Temperature (IQR)a 25 °C 30 °C Spring (n = 25) 59.2 (38.4–87.5) 2.8 (0.0–22.1) Summer (n = 18) 94.9 (69.6–98.4) 21.3 (13.6–48.8) Autumn (n = 31) 58.1 (32.1–90.0) 6.6 (0.0–23.5) Winter (n = 41) 40.6 (26.2–77.1) 5.9 (1.0–29.8) pb <0.0001 <0.05 Season Median % Readings Higher Than Indicated Temperature (IQR)a 25 °C 30 °C Spring (n = 25) 59.2 (38.4–87.5) 2.8 (0.0–22.1) Summer (n = 18) 94.9 (69.6–98.4) 21.3 (13.6–48.8) Autumn (n = 31) 58.1 (32.1–90.0) 6.6 (0.0–23.5) Winter (n = 41) 40.6 (26.2–77.1) 5.9 (1.0–29.8) pb <0.0001 <0.05 a IQR = interquartile range. The median (IQR) percentage of readings higher than 35 °C was 0.00% (0.00–0.00%) for all seasons (p = 0.858). b Kruskal-Wallis test. Open in new tab Table 2 Percentage of Infuser Temperature Readings Exceeding 25 and 30 °C Across the Seasons Season Median % Readings Higher Than Indicated Temperature (IQR)a 25 °C 30 °C Spring (n = 25) 59.2 (38.4–87.5) 2.8 (0.0–22.1) Summer (n = 18) 94.9 (69.6–98.4) 21.3 (13.6–48.8) Autumn (n = 31) 58.1 (32.1–90.0) 6.6 (0.0–23.5) Winter (n = 41) 40.6 (26.2–77.1) 5.9 (1.0–29.8) pb <0.0001 <0.05 Season Median % Readings Higher Than Indicated Temperature (IQR)a 25 °C 30 °C Spring (n = 25) 59.2 (38.4–87.5) 2.8 (0.0–22.1) Summer (n = 18) 94.9 (69.6–98.4) 21.3 (13.6–48.8) Autumn (n = 31) 58.1 (32.1–90.0) 6.6 (0.0–23.5) Winter (n = 41) 40.6 (26.2–77.1) 5.9 (1.0–29.8) pb <0.0001 <0.05 a IQR = interquartile range. The median (IQR) percentage of readings higher than 35 °C was 0.00% (0.00–0.00%) for all seasons (p = 0.858). b Kruskal-Wallis test. Open in new tab Effects of daytime storage location. Participants were stratified into 4 subgroups based on where the infuser was carried during daytime use to investigate the effect of daytime storage location on infuser temperature (Table 3). Table 3 Percentage of Infusor Temperature Readings Exceeding 25 and 30 °C Across Various Storage Methods Method Median % Readings Higher Than Indicated Temperature (IQR)a 25 °C 30 °C Daytime  Carried on body (n = 86) 80.3 (47.8–90.7) 7.7 (0.0–25.8)  Hung on body (n = 9) 69.5 (60.5–95.8) 0.6 (0.0–2.5)  Carried off body (n = 8) 84.9 (58.7–94.8) 6.8 (0.5–19.5)  Hung off body (n = 5) 83.2 (21.2–99.0) 4.4 (0.0–31.9)  All participants (n = 115) 81.0 (47.8–92.0) 7.1 (0.0–26.0)  pb 0.915 0.387 Nighttime  On body or in bed (n = 22) 100.0 (79.3–100.0) 65.0 (13.2–86.5)  On bed (n = 55) 17.7 (1.8–100.0) 0.0 (0.0–19.6)  Away from bed (n = 27) 5.8 (0.7–17.6) 0.0 (0.0–0.0)  All participants (n = 112) 36.1 (2.6–100.0) 0.0 (0.0–38.6)  pb <0.0001 <0.0001 Method Median % Readings Higher Than Indicated Temperature (IQR)a 25 °C 30 °C Daytime  Carried on body (n = 86) 80.3 (47.8–90.7) 7.7 (0.0–25.8)  Hung on body (n = 9) 69.5 (60.5–95.8) 0.6 (0.0–2.5)  Carried off body (n = 8) 84.9 (58.7–94.8) 6.8 (0.5–19.5)  Hung off body (n = 5) 83.2 (21.2–99.0) 4.4 (0.0–31.9)  All participants (n = 115) 81.0 (47.8–92.0) 7.1 (0.0–26.0)  pb 0.915 0.387 Nighttime  On body or in bed (n = 22) 100.0 (79.3–100.0) 65.0 (13.2–86.5)  On bed (n = 55) 17.7 (1.8–100.0) 0.0 (0.0–19.6)  Away from bed (n = 27) 5.8 (0.7–17.6) 0.0 (0.0–0.0)  All participants (n = 112) 36.1 (2.6–100.0) 0.0 (0.0–38.6)  pb <0.0001 <0.0001 a IQR = interquartile range. Patients who did not specify a storage method (7 for daytime and 8 for nighttime) were not included in this analysis. Daytime storage location analyses were conducted using daytime temperature readings only. Nighttime storage location analyses were conducted using nighttime temperature readings only. For both daytime and nighttime storage analyses, the median (IQR) percentage of readings that were higher than 35 °C was 0.00% (0.00–0.00%) for all storage locations (daytime storage locations, p = 0.634; nighttime storage locations, p = 0.0005). b Kruskal-Wallis test. Open in new tab Table 3 Percentage of Infusor Temperature Readings Exceeding 25 and 30 °C Across Various Storage Methods Method Median % Readings Higher Than Indicated Temperature (IQR)a 25 °C 30 °C Daytime  Carried on body (n = 86) 80.3 (47.8–90.7) 7.7 (0.0–25.8)  Hung on body (n = 9) 69.5 (60.5–95.8) 0.6 (0.0–2.5)  Carried off body (n = 8) 84.9 (58.7–94.8) 6.8 (0.5–19.5)  Hung off body (n = 5) 83.2 (21.2–99.0) 4.4 (0.0–31.9)  All participants (n = 115) 81.0 (47.8–92.0) 7.1 (0.0–26.0)  pb 0.915 0.387 Nighttime  On body or in bed (n = 22) 100.0 (79.3–100.0) 65.0 (13.2–86.5)  On bed (n = 55) 17.7 (1.8–100.0) 0.0 (0.0–19.6)  Away from bed (n = 27) 5.8 (0.7–17.6) 0.0 (0.0–0.0)  All participants (n = 112) 36.1 (2.6–100.0) 0.0 (0.0–38.6)  pb <0.0001 <0.0001 Method Median % Readings Higher Than Indicated Temperature (IQR)a 25 °C 30 °C Daytime  Carried on body (n = 86) 80.3 (47.8–90.7) 7.7 (0.0–25.8)  Hung on body (n = 9) 69.5 (60.5–95.8) 0.6 (0.0–2.5)  Carried off body (n = 8) 84.9 (58.7–94.8) 6.8 (0.5–19.5)  Hung off body (n = 5) 83.2 (21.2–99.0) 4.4 (0.0–31.9)  All participants (n = 115) 81.0 (47.8–92.0) 7.1 (0.0–26.0)  pb 0.915 0.387 Nighttime  On body or in bed (n = 22) 100.0 (79.3–100.0) 65.0 (13.2–86.5)  On bed (n = 55) 17.7 (1.8–100.0) 0.0 (0.0–19.6)  Away from bed (n = 27) 5.8 (0.7–17.6) 0.0 (0.0–0.0)  All participants (n = 112) 36.1 (2.6–100.0) 0.0 (0.0–38.6)  pb <0.0001 <0.0001 a IQR = interquartile range. Patients who did not specify a storage method (7 for daytime and 8 for nighttime) were not included in this analysis. Daytime storage location analyses were conducted using daytime temperature readings only. Nighttime storage location analyses were conducted using nighttime temperature readings only. For both daytime and nighttime storage analyses, the median (IQR) percentage of readings that were higher than 35 °C was 0.00% (0.00–0.00%) for all storage locations (daytime storage locations, p = 0.634; nighttime storage locations, p = 0.0005). b Kruskal-Wallis test. Open in new tab No significant differences were noted among the locations in the proportion of median temperature readings of >25 °C (p = 0.915), >30 °C (p = 0.387), or >35 °C (p = 0.634) (Table 3). Effect of bedtime storage location. Participants were stratified into 3 subgroups based on where the infuser was stored while the participant was sleeping to investigate the effect of overnight storage location on infuser temperature. The mean temperature readings at bedtime and at 5-minute intervals after bedtime are presented in Figure 6 for each overnight storage location. A significant difference in the mean temperature profiles was noted across the 3 groups (GLMM analysis, p < 0.001), with infusers stored “on body or in bed” having higher mean temperatures than infusers stored “on bed” or “away from bed.” The percentages of bedtime temperature readings that were >25, >30, and >35 °C for the various overnight storage locations are summarized in Table 3. Significant differences in the median percentage of bedtime temperature readings of >25 °C (p < 0.0001), >30 °C (p < 0.0001), and >35 °C (p < 0.001) were found when the storage locations were compared. Figure 6 Open in new tabDownload slide Mean infuser temperatures at 5-minute intervals from 2 hours before bedtime (−120 to 0 minutes) to 8 hours after bedtime (0 to 480 minutes). Data are grouped on the basis of where the infuser was stored overnight. Figure 6 Open in new tabDownload slide Mean infuser temperatures at 5-minute intervals from 2 hours before bedtime (−120 to 0 minutes) to 8 hours after bedtime (0 to 480 minutes). Data are grouped on the basis of where the infuser was stored overnight. Post hoc analyses indicated that the median percentage of bedtime readings of >25 °C was significantly higher for participants who stored their infusers “on body or in bed” (100.0%) compared with those who stored their infuser “on bed” (17.7%, p < 0.001) or “away from bed” (5.8%, p < 0.001). A significant difference in the median percentage of bedtime readings of >25 °C was also detected between participants storing their infusers “on bed” and “away from bed” (p < 0.05). The median percentage of bedtime readings of >30 °C was also significantly higher for participants who stored their infuser “on body or in bed” (65.0%) compared with those who stored their infuser “on bed” (0.0%, p < 0.001) or “away from bed” (0.0%, p < 0.001) and again was higher for participants storing their infuser “on bed” than for those storing their infuser “away from bed” (p < 0.001). Only 4 participants recorded infuser temperatures of >35 °C during the bedtime period. All 4 stored their infuser “on body or in bed.” When all temperature readings recorded during the bedtime period were pooled and averaged, the mean ± S.D. bedtime temperature for participants who stored the infuser “on body or in bed” was 29.9 ± 3.5 °C. This was much higher than the mean ± S.D. bedtime readings calculated for participants storing their infusers “on bed” (23.5 ± 4.7 °C) or “away from bed” (21.1 ± 4.7 °C). This trend was consistent with the findings presented in Table 3 and evident when the analyses were extended to include all temperature readings collected during the 24-hour infusion period (mean ± S.D. for “on body or in bed,” 28.0 ± 3.6 °C; for “on bed,” 25.2 ± 4.3 °C; for “away from bed,” 23.8 ± 4.8 °C). Discussion To our knowledge, this was the first study to measure the temperature of i.v. infusion devices within a large population of HITH patients, spanning all seasons over a 12-month period. The findings revealed significant predictable fluctuations in the temperature of the infuser throughout a 24-hour period in HITH patients, suggesting that the current use of a steady-state temperature model for testing the stability of antibiotic solutions is not reflective of clinical practice. Further, this approach may not be appropriate for determining antibiotic stability or suitability in home-based care, particularly for antimicrobial agents that are sensitive to temperature-related degradation. It is possible that the temperature fluctuations found in this study may be a significant independent variable in the overall assessment of drug stability in this situation, along with other recognized variables such as the drug itself, its concentration, and the diluent. However, given that ambient temperature varies according to the geography and season of each health institution, it is unrealistic to develop a stability testing model to suit all HITH scenarios. As such, the concept of a validated “mean kinetic temperature range,” commonly used in the storage and transportation of perishable goods and commercially manufactured pharmaceuticals, is of interest.15 Published antibiotic stability studies using a fixed room temperature of 37 °C have shown acceptable antibiotic degradation rates (i.e., <10% over 24 hours) for some i.v. antibiotics in solution, including vancomycin,6 ticarcillin–clavulanic acid,7,8 piperacillin–tazobactam,7,8 cefazolin,10 buffered flucloxacillin,11 and buffered penicillin G sodium.9 Consequently, these drugs are generally deemed appropriate for use in HITH programs for continuous i.v. administration over 24 hours. This high temperature was chosen to account for warmer-than-expected ambient temperatures as well as the possible effect of body heat on infusion devices or the antibiotic solution. However, it is not clear whether i.v. antibiotics that degrade at unacceptable rates (i.e., >10%) at lower ambient temperatures can be safely and effectively used in HITH programs. For example, meropenem and unbuffered penicillin G sodium exhibit unacceptable degradation rates when exposed to a constant temperature of 25–26 °C over 24 hours, with meropenem showing greater than 10% degradation after 12 hours at clinically significant concentrations in one study12 and unbuffered penicillin G sodium degrading after approximately 13 hours in another.4 While ceftazidime has been shown to be stable over 24 hours at a fixed room temperature of 25 °C, one study deems it unstable at 37 °C with an unsuitable degradation rate occurring after 8 hours.5 Unbuffered flucloxacillin also degrades at unacceptable rates when exposed to 37 °C for 7 hours within a 24-hour period (where it was exposed to 31 °C for the remaining time).16 Unfortunately, none of these stability testing conditions mimic the varying temperatures noticed in real-life practice, as has been reviewed in this study. Thus the antibiotics that degrade at lower fixed room temperatures are not used with confidence over longer infusion times in HITH programs, or a buffer is added to improve stability at higher temperatures. If future stability studies utilize a varying ambient temperature model such as is broadly described in this report, the suitability of such i.v. antibiotics in HITH programs may be better realized. Within our study, where the antibiotic infuser was carried during the day did not significantly impact infuser temperatures. In contrast, the overnight storage location made a marked difference to the mean temperature and to the proportion of elevated temperature readings observed. Significantly cooler temperatures were recorded when participants stored their infusion device “away from bed” while sleeping, with under 20% of bedtime readings being greater than 25 °C and less than 3% being over 30 °C when this storage location was used. Importantly, the observed mean temperature of infusions over the full 24 hours, where the infuser was optimally stored overnight (away from bed), was 23.9 °C. This is lower than the temperatures often used to measure the stability of antibiotic infusions in the literature (25–37 °C).5–12 The findings suggest that professionals caring for patients receiving i.v. antibiotic infusions at home should recommend several strategies for lowering infuser temperatures, such as (1) keeping the infuser away from the bed at nighttime rather than on the bed or on the body and (2) ensuring that infusers are kept cooler during summer months. These efforts should reduce the degradation of the antibacterial and allow higher antibiotic doses to remain in solution, which is an important concept of antimicrobial stewardship and the prevention of antibiotic resistance.17 This study had several limitations. First, the data logger thermometer was attached to the end of the infuser and was not immersed within the antibiotic solution. Consequently, the actual temperature of the antibiotic solution was not measured and may have varied from the reported temperature. Second, the double-walled structure of the infusion device, which incorporates an air gap between the inner drug reservoir and the outer wall of the device, may have insulated the antibiotic solution and dampened temperature fluctuation. Third, the position of the data logger at the end of the infuser may have also limited the ability of the thermometer to register the influence of body heat, particularly for individuals who carried the infuser on their body. For these people, the section of the infuser adjacent to the body may have been exposed to higher temperatures than were detected by the thermometer. Finally, the study was conducted at a single site, and there may be unknown variables within this site that affected the outcome. Further work is required to validate these findings. If valid, these findings suggest that HITH programs should reassess antibiotic stability testing in light of the temperature fluctuations described herein. While future stability testing can still be conducted in vitro, we recommend that the prepared solutions are exposed to the temperature fluctuations broadly described in this report to be more representative of clinical practice. Conclusion Antibiotics administered to HITH patients via continuous infusion were frequently exposed to temperatures in excess of 25 °C. Specific patient behaviors and seasonal and chronological factors influenced temperatures. The findings challenge the validity of current fixed-temperature models for testing stability, which do not reflect conditions found in clinical practice. Disclosures The authors have declared no potential conflicts of interest. Acknowledgments The research assistance of Peter Mastello, B.Pharm., and Tara Prowse, B.Pharm., M.Clin.Pharm., is acknowledged. Catherine Paavola, RN, Janelle Sawers, RN, and the nursing staff within the Acute Post-Acute Care service are acknowledged for their assistance related to data collection. References 1 Trissel LA . Handbook on injectable drugs . 18th ed. Bethesda, MD : Oxford University Press ; 2013 : ix . Google Preview WorldCat COPAC 2 Outpatient parenteral antimicrobial therapy . In: Therapeutic guidelines: antibiotic. Version 15 . Melbourne, Australia : Therapeutic Guidelines Limited ; 2014 : 549 . WorldCat COPAC 3 Poole SG Dooley MJ . Drugs in ambulatory infusion devices. The appropriate temperature for stability testing . Aust J Hosp Pharm . 1999 ; 29 : 328 – 9 . Google Scholar Crossref Search ADS WorldCat 4 Vella-Brincat JW Begg EJ Gallagher K . Stability of benzylpenicillin during continuous home intravenous therapy . J Antimicrob Chemother . 2004 ; 53 : 675 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Viaene E Chanteux H Servais H . Comparative stability studies of antipseudomonal beta-lactams for potential administration through portable elastomeric pumps (home therapy for cystic fibrosis patients) and motor-operated syringes (intensive care units) . Antimicrob Agents Chemother . 2002 ; 46 : 2327 – 32 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Raverdy V Ampe E Hecq JD . Stability and compatibility of vancomycin for administration by continuous infusion . J Antimicrob Chemother . 2013 ; 68 : 1179 – 82 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Arlicot N Rochefort GY Schlecht D . Stability of antibiotics in portable pumps used for bronchial superinfection: guidelines for prescribers . Pediatrics . 2007 ; 120 : 1255 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Prescott WA Jr Gentile AE Nagel JL Pettit RS . Continuous-infusion antipseudomonal beta-lactam therapy in patients with cystic fibrosis . P T . 2011 ; 36 : 723 – 6 . Google Scholar PubMed WorldCat 9 MacDougal DA McWhinney BC . Stability of buffered benzylpenicillin solutions for outpatient parenteral antimicrobial therapy . J Pharm Pract Res . 2014 ; 44 : 26 – 8 . Google Scholar Crossref Search ADS WorldCat 10 Stiles ML Tu YH Allen LV Jr . Stability of cefazolin sodium, cefoxitin sodium, ceftazidime, and penicillin G sodium in portable pump reservoirs . Am J Hosp Pharm . 1989 ; 46 : 1408 – 12 . Google Scholar PubMed WorldCat 11 To TP Ching MS Ellia AG . Stability of intravenous flucloxacillin solutions used for hospital-in-the-home . J Pharm Pract Res . 2010 ; 40 : 101 – 5 . Google Scholar Crossref Search ADS WorldCat 12 Berthoin K Le Duff CS Marchand-Brynaert J . Stability of meropenem and doripenem solutions for administration by continuous infusion . J Antimicrob Chemother . 2010 ; 65 : 1073 – 5 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Australian Bureau of Statistics . Canberra: census of population and housing: nature and content ( 2011 ). www.ausstats.abs.gov.au/ausstats/subscriber.nsf/0/410FDD3C31C0BE1FCA257673001BFC61/$File/20080_2011.pdf (accessed 2015 Aug 23). 14 Urbaniak GC Plous S . Research randomizer . www.randomizer.org (accessed 2015 Aug 23). 15 Pharmaceutical stability (general chapter 1150) . In: The United States pharmacopeia, 29th rev., and The national formulary , 24th ed. www.pharmacopeia.cn/v29240/usp29nf24s0_c1150.html (accessed 2016 May 12). 16 Carroll JA . Stability of flucloxacillin in elastomeric infusion devices . J Pharm Pract Res . 2005 ; 35 : 90 – 3 . Google Scholar Crossref Search ADS WorldCat 17 Roberts JA Kruger P Paterson DL . Antibiotic resistance—what’s dosing got to do with it? Crit Care Med 2008 ; 36 : 2433 – 40 . Google Scholar Crossref Search ADS PubMed WorldCat Copyright © 2017 by the American Society of Health-System Pharmacists, Inc. All rights reserved.
TI - Temperature profiles of antibiotic-containing elastomeric infusion devices used by ambulatory care patients
JF - American Journal of Health-System Pharmacy
DO - 10.2146/ajhp151071
DA - 2017-07-01
UR - https://www.deepdyve.com/lp/oxford-university-press/temperature-profiles-of-antibiotic-containing-elastomeric-infusion-cwcV2EA5nb
SP - 992
VL - 74
IS - 13
DP - DeepDyve
ER -