Article Text

Download PDFPDF

Systematic review of stability data pertaining to selected antibiotics used for extended infusions in outpatient parenteral antimicrobial therapy (OPAT) at standard room temperature and in warmer climates
  1. Stephen John Perks1,2,
  2. Cassie Lanskey2,
  3. Niechole Robinson2,
  4. Tilley Pain1,2,
  5. Richard Franklin1
  1. 1 School of Public Health, Medicine and Vetinary Sciences, James Cook University, Townsville, Queensland, Australia
  2. 2 Pharmacy, Townsville Hospital and Health Service, Townsville, Queensland, Australia
  1. Correspondence to Mr Stephen John Perks,School of Public Health, Medicine and Vetinary Sciences, James Cook University, Townsville, Queensland, Australia; steve.perks{at}


Aim To determine if there are sufficient stability data to confirm appropriate prescribing of antibiotics commonly used in outpatient parenteral antimicrobial therapy (OPAT) in warmer climates.

Data sources Four databases were systematically searched using the terms ‘beta-lactams’, or ‘antibiotics’, or ‘anti-bacterial agents’ and ‘drug stability’ or drug storage’ for studies specific to drug stability published between 1966 and February 2018.

Study selection The search strategy initially identified 2879 potential articles. After title and abstract review, the full-texts of 137 potential articles were assessed, with 46 articles matching the inclusion and exclusion criteria included in this review.

Results A large volume of stability data is available for the selected drugs. Stability data at temperatures higher than 25°C were available for several of the medications, however few drugs demonstrated stability in warmer climates of 34°C or higher. Only buffered benzylpenicillin, cefoxitin and buffered flucloxacillin were found to have stability data supporting OPAT in warmer climates. Sequential data, profiling the drug for an extended period in solution under refrigeration prior to the run-out period at the higher temperatures, are also lacking.

Limitations This study was limited by including only peer reviewed articles. There may be further grey literature supporting the stability of some of the drugs mentioned.

Conclusion There are insufficient stability data of antibiotic use in warmer climates. Studies to verify the stability and appropriate use of many antibiotics used in OPAT at standard room temperature and in warmer climates are urgently required. Several drugs in current use in the OPAT settings are lacking stability data.

Implications Further research in this field is needed to develop structured evidence-based guidelines. Results of this review should be further compared with observed patient outcomes in current clinical practice.

  • hospital in the home
  • outpatient parenteral antimicrobial therapy
  • antibiotic administration
  • stability

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Outpatient Parenteral Antimicrobial Therapy (OPAT) is one part of the overarching Hospital in The Home programme. OPAT patients can receive highly concentrated intravenous antibiotics via portable infusion pumps over repeated 24 hours infusions after discharge from hospital. Medication stability is a critical factor when deciding if these antibiotics are appropriate for use as extended infusions.1 Poor drug stability may result in the patient receiving less than the intended dose required to treat the infection2–6 leading to extended treatment durations, treatment failure, readmissions to hospital and antibiotic resistance.7 Therefore, only antibiotics with proven stability applicable to OPAT conditions should be used in this setting.

Stability data must be specific to the conditions of use and the mode of therapy. Conditions affecting stability include: the dose/concentration of the therapy; length of time the antibiotic is in solution; diluents used; additives in solution; and temperature of solution during storage and administration. Most stability data include these parameters up to room temperature (25°C), but not higher temperatures. Room temperature studies are generally standardised at 20–25°C.1 8 Stability data are lacking for many antibiotics in solution above room temperature.9 In warmer climates, the solution temperature may exceed 25°C for over 75% of the 24 hours run-out period, occasionally reaching temperatures of up to 38°C.10 High concentration OPAT antibiotics are prepared in solution and often refrigerated up to 7 days prior to the 24 hours run-out period. Therefore, to be accurate, stability data must include the entire time profile of the antibiotic in solution, as the degradation can start even during refrigeration. A recent study mapping the temperature within the solution in these devices over the 24 hours run-out period in a warmer climate recommended stability data be obtained at 34°C following exposure of up to 7 days of refrigeration.11

A recent systematic review published in 2017 found that there were no published studies that complied completely with UK national standards specific to OPAT.9 The UK standards are detailed and describe the minimum data required from a study to be applicable for use in this setting. Recommendations from the review were to promote further research and publications pertaining to antimicrobial stability data to support OPAT. Current OPAT services continue to run without adequate guidance with regard to the stability and suitability of medications in this setting. What we propose is to review all literature to determine if there are any stability data available for the drugs used in this setting. This will give some level of guidance to clinicians currently working in the OPAT setting and will also give researchers a starting platform for further stability and suitability studies for this setting.

Our aims were to (1) To perform a systematic review of antibiotic stability data literature. (2) To determine if there are sufficient data to confirm >90% stability for commonly used, or historically used, antibiotics in OPAT in standard (20–25oC) and warmer climates (34°C or above).


This systematic review follows PRISMA (Preferred Reporting for Systematic reviews and Meta-Analyses) guidelines.12 This protocol was not registered.

Search strategy and study eligibility

Original research articles written in English language and published from 1966 to 1 February 2018 were eligible for inclusion. Databases searched included MEDLINE (1966–2018), EMBASE (all years), CINAHL (all years) and the Cochrane library.

Searches were performed by four of the authors all of which are currently employed as clinical pharmacists. Due to the subtle differences in the database’s cataloguing the following search terms were employed: MEDLINE and EMBASE: beta-lactams/ AND (Drug Stability/or Drug Storage/); CINAHL: Antibiotics AND (Drug Stability or Drug Storage); Cochrane: Anti-Bacterial Agents AND (Drug Stability or Drug Storage) (figure 1).

Figure 1

Systematic review search method.

Titles were scanned and abstracts reviewed to identify articles relevant to stability. Drugs in this review were historically used in OPAT and included: amoxycillin, ampicillin, benzylpenicillin, cefepime, cefotaxime, cefoxitin, ceftazidime, doripenem, ertapenem, flucloxacillin, meropenem, piperacillin (and tazobactam), and ticarcillin (and clavulanate). Studies investigating these drugs in solutions at a specified temperature, and recording an outcome reflecting final stability after exposure were included. The abstracts were then read to identify suitable articles for full-text review.

Excluded papers included studies that: didn't include the listed drugs; only focused on validating high performance liquid chromatography (HPLC) analytical methodologies; investigated drug stability to external agents such as β-lactamases or renal dyhydropeptidases; were performed in plasma/blood/urine/serum/milk or animals or renal dialysate; evaluated admixtures of multiple drugs in the one solution; assessed previously frozen samples unless there was clear evaluation of stability in the liquid state also; involved microwave thawing; and investigations of pharmacokinetics on drug complexation within cyclodextrins. Microbial studies were also excluded as only physicochemical stability data in solution are of interest. Studies using iodometric analysis rather than HPLC analysis were also excluded as there were sufficient high-quality papers using HPLC as the analysis method. As we wanted data to compare to OPAT conditions, studies were excluded if they were of insufficient duration or temperature for comparison. Therefore, studies were excluded if they did not contain results at a minimum of 20°C and have data to support at least 12 hours of stability during the run-out phase. Studies that tested concentrations of a subclinical dose after preparation in a 48 mL−240 mL infusor were also excluded.

Data extraction and quality appraisal

Data extracted from the eligible articles included author details, year of publication, the drug and concentration studied, diluent used, buffers or other additives in solution, duration of exposure and temperature, and stability demonstrated under each condition. Stability was defined for this study as ≥90%1 of drug concentration by the end of the period. It is important to note that this 90% or greater standard is different to the UK standard,13 but is in line with the consensus of the majority of stability studies performed around the world. Studies investigating multiple temperatures were included if sequential exposures occurred. Sequential exposure refers to samples exposed to a refrigerated temperature for a period followed by a period at a higher temperature. All included articles were peer reviewed.

Data extracted for discussion were grouped into standard ‘room temperature’ and ‘warmer climate’. Standard room temperature included temperatures as low as 20°C, and warmer climate included data studied at 34°C or above. From a recent study 34°C was chosen, mapping solution temperature within OPAT devices in a warmer climate.11

The specific inclusion and exclusion criteria acted to ensure only stability studies that would be applicable to the OPAT setting were included in the results. The quality of studies performed was mentioned in the discussions if noted to be poor, or if individual results appeared to be inconsistent within the study or compared with the majority of other results.

We proposed chemical stability for extended infusions in standard and warmer climates should meet specific criteria. First, stability should be demonstrated at a minimum of 20–25oC8 for ‘standard room temperature’ or 34°C or above for ‘warmer climates’, for the 24 hours run-out period.11 Second, the concentration should reflect the clinical dose in a 48–240 mL infusor as these are the most common volumes of home infusors. Third, stability should be assessed by sequential refrigerated storage followed by the run-out period, to reflect the batch manufacturing and ordering of these devices.


The initial search found 2879 papers after duplicates were removed. The most common reasons for exclusion were papers that were developing analysis methodologies, or investigating analysis and stability of drugs to external agents or within tissue/milk/urine samples. Following preliminary title and abstract review, 137 papers remained for full-text review. Additional papers were excluded because: 18 full-texts could not be obtained; six were letters to the editor or not peer reviewed; 18 did not meet the analysis methodology requirements of the review; 26 did not have specific duration or temperature of exposure data that would be useful in the OPAT setting; 10 were studies on drug degradation kinetics; 4 did not have concentrations high enough to be useful in the OPAT setting; and 9 were not applicable as they were not investigating the named drugs or they were investigating them in combination with other drugs in solution. Forty-six articles were included in the systematic review.

The papers could be broken into four clinical groups with some papers overlapping into multiple groups. The groups were penicillins (n=13), cephalosporins (n=23), carbapenems (n=7) and combination antipseudomonal penicillins (n=7) as summarised in tables 1–4, respectively. These tables list the highest temperature and timeframes in which stability was successfully demonstrated. Stability conditions considered appropriate for OPAT were defined as run-out temperatures at or above 20°C for ‘room temperature’ studies and at or above 34°C for warmer temperature studies. A minimum of 12 hours of stability in the run-out period was needed to be useful for discussion when comparing to the OPAT setting. Each line of tables 1–4 represents the parameters of investigation for a specific study that verified stability of greater than 90%. Due to the large amount of study conditions showing instability, parameters of investigations that confirmed the instability of the drugs were not included.

Table 1

Experimental conditions verifying stability of selected penicillins

Table 2

Experimental conditions verifying stability of selected cephalosporins

Table 3

Experimental conditions verifying stability of selected carbapenems

Table 4

Experimental conditions verifying stability of selected antipseudomonal penicillins


Medications used in extended infusions must demonstrate stability at ‘room temperature’ regardless of the climatic conditions where the infusion occurs. Historically, room temperature is described as 25°C and is the temperature in which most stability studies are performed. We propose chemical stability for extended infusions in OPAT should meet three specific criteria. First, stability should be demonstrated at 20°C–25°C8 for standard room temperature or at 34°C or above for warmer climate for the 24 hours run-out period.11 Second, the concentration should reflect the clinical dose in 240 mL as it is the most common volume of home infusors. Third, stability data must be assessed by sequential refrigerated storage followed by the run-out period, to reflect the batch manufacturing and ordering of these devices. If stability is not studied sequentially it does not reflect OPAT infusion solution conditions, since the infusors can be stored refrigerated for up to 7 days prior to exposure to the 24 hours run-out period at the higher temperature. Based on these criteria, this systematic review demonstrates some antibiotics may be unsuitable for home infusion devices in normal ‘room temperature’ or in warmer climates in the OPAT setting.

Amoxicillin had no studies using sequential data and therefore may be inappropriate for use in OPAT. Ampicillin had one sequential study using a sodium phosphate buffer demonstrating stability for 24 hours at 25°C after 48 hours refrigeration.14 However, this was for a short refrigeration period, and was at a concentration that would be considered subtherapeutic for most patients. Therefore, further studies are required to verify ampicillin stability at higher doses, and for longer refrigeration periods before recommending it for use in the OPAT setting. Benzylpenicillin and flucloxacillin were originally deemed inappropriate for use in OPAT, but with the introduction of buffering methods both have proven a viable choice for therapy.15 16 Unbuffered flucloxacillin has sequential data supporting stability at ‘room temperature’,17 but only buffered flucloxacillin demonstrated stability at higher temperatures appropriate for a warmer climate.

Studies on cephalosporins show some may not be suitable for use in OPAT settings in warmer climates.

No studies on cefepime met all three of our nominated criteria for testing stability in warmer climates. Sequential data for cefepime showed stability after 7 days refrigeration and 24 hours at room temperature of 22°C–24°C18 and up to 29°C.19 However, there were no data showing stability in warmer climates, and only non-sequential data demonstrating stability up to 13 hours at 37°C.20 Cefotaxime had several studies showing stability at ‘room temperature’, but no data for higher temperatures nor sequential data. Therefore, cefotaxime is inappropriate for use in OPAT based on our criteria because of lack of evidence. Cefoxitin sequential data showed stability for 30 days at −20°C, then 4 days at 5°C, then 1 day at 37°C.21 Freezing then refrigerating may reduce degradation of the solution as the temperature of the device may be lower than the refrigerated temperature during the thawing phase. Devices are not routinely frozen, so the applicability to the OPAT situation is unknown. Although it is likely to be stable in all OPAT conditions, based on our three criteria, we suggest further investigations are needed to verify the stability of cefoxitin for OPAT at higher temperatures.

Interpretation of ceftazidime data (table 2) is complex because of its degradation limits, plus limits placed on the build-up of the toxic degradation by-product pyridine. This review will initially review ceftazidime stability on its own followed by the more complicated interpretation of the pyridine build-up.

Ceftazidime at 60 mg/mL solution was stable after sequential refrigeration and run-out at 27°C.22 However, inconsistencies within that study are evident as both concentrations show improved stability over the longer refrigeration period—an unexpected outcome as longer refrigeration would be associated with higher degradation. Stiles et al state a 36.6 mg/mL solution is stable after exposure for 30 days at −20°C, then 4 days at 5°C, then 1 day at 37°C. Nahata et al showed sequential data of a 100 mg/mL solution of ceftazidime with a run-out temperature of 22°C. Therefore, some ceftazidime stability studies meet the three criteria but suitability of ceftazidime use in OPAT setting is complicated by the breakdown product pyridine.

Ceftazidime releases pyridine from the pyridinium moiety on the C3 side chain.23 Pyridine build-up is toxic with evidence in animal models of its carcinogenic effects.24 However, there is inadequate evidence for the carcinogenicity of pyridine in humans.24 Known effects include dizziness, headache, nausea, central nervous system issues, anorexia, abdominal pain and pulmonary congestion.24 Previously, pyridine in ceftazidime solutions was limited to 1.1 mg/mL by the United States Pharmacopeia (USP).25 However, the USP have now listed it as a class II solvent with a permitted daily exposure (PDE) of 2 mg.26 This new pyridine threshold aligns with the PDE that has been listed by the European Medicines Agency (EMA)27 since the year 2000. The British Pharmacopoeia (BP), lists pyridine limits at 0.5% w/v. However, the BP does not specify to which volume it is referring (ie, original phial size or volume of infusion). The lack of clarity around volume will impact how much pyridine a patient receives. For example, if three patients receive a dose of 6 g ceftazidime with volumes of a 30 mL push, a 240 mL OPAT infusion and a 1000 mL inpatient infusion, the acceptable amounts of pyridine in solutions using the BP limit equates to 150 mg, 1200 mg and 5000 mg, respectively. This reasoning led us to choose the USP and EMA standards to limit the PDE of pyridine. We recommend ceftazidime stability studies have concurrent analysis of pyridine levels.

No studies in this review recorded sequential ceftazidime stability data concurrently with pyridine levels (Bednar et al,22 Stiles et al,21 Nahata et al 28). Furthermore, all sequential studies in this review may breach the previous pyridine standard of 1.1 mg/mL at higher temperatures when compared with the results of Viaene et al,20 which shows solutions from 83 mg/mL to 120 mg/mL will breach the 1.1 mg/mL threshold by a factor of 2–3 at 16 hours when stored at 37°C. Despite the lower concentration of ceftazidime at 36 mg/mL the Stiles21 study is likely to breach the acceptable 1.1 mg/mL pyridine levels by 24 hours given the longer time frame (30 days at −20°C, 4 days under refrigeration followed by 24 hours at 37°C).

The introduction of the 2 mg PDE26 27 standard for pyridine exposure means the old standard of the 1.1 mg/mL25 should be disregarded. Bourget et al,29 measured the total delivered amount of pyridine over various full-day infusions comparing two 6 g infusion over 11.5 hours each for a total of 23 hours of run-out and a single 23 hours infusion of 12 g. This split method (2×6 g infusions) was tested to try to reduce the total daily exposure of pyridine. Although it was a successful reduction, the amount of pyridine the patient still received was 26.4 mg and 30.4 mg when run-out at 22°C and at 33°C, respectively. This is 13–15 times the acceptable threshold. Using the 12 g infusion over 23 hours run-out at 33°C would deliver 91.5 mg of pyridine (45 times the PDE in USP and EMA) to the patient. The Bourget study did not record sequential data we recommend for stability studies in OPAT settings which suggests the pyridine could be even higher. Plasse et al 30 investigated a lower dose of 2 g daily which would ordinarily be considered a subtherapeutic dose as most adults would be on a 6 g daily dose. The 2 g infusion over 24 hours at room temperature delivered a total dose of pyridine of 4.56 mg. Even at this lower dose of 2 g daily and at room temperature the threshold is more than doubled over the period. These results clearly show the rationale for concurrently measuring pyridine is compelling. Furthermore, the lack of evidence of the effect of pyridine in humans implies further investigations are urgently required to guide practitioners.

Results in table 3 show stability data for carbapenems do not meet our three criteria. Stability up to 12–16 hours was demonstrated, with two studies showing stability of doripenem for 24 hours.31 32 However, neither study reported sequential data after refrigeration and are therefore not useful in the OPAT setting.

Table 4 shows results for antipseudomonal penicillins. Piperacillin/tazobactam has high stability up to 28 days refrigeration followed by 3 days at 23°C33 but only if buffered. Non-sequential data show stability of piperacillin/tazobactam at 25°C for 21 days.34 However, there are no sequential stability data for antipseudomonal penicillins at higher temperatures. Results from the two prementioned studies suggest stability is likely under the higher temperature tested conditions. However, further studies are needed to confirm this. Ticarcillin/clavulanate stability is tested in one non-sequential study showing stability for 24 hours at 35°C.3 Given the lack of evidence under specific OPAT conditions, it may be more suitable to use an alternative antibiotic in this setting.

This review highlights areas of therapy currently occurring in many OPAT centres that may not be providing the best possible patient care due to potential patient underdosing. This could have severe patient outcomes which should be investigated further. This review also poses the question of ‘what affect this potential under-dosing would have on antimicrobial resistance patterns’. Giving inappropriate doses of antimicrobials is linked with resistance and the emergence of multiresistant organisms. This issue should be further investigated from an antimicrobial stewardship point of view.

Table 5 has been generated as a quick stability reference for clinicians. The table lists antibiotics which have stability data after refrigeration and exposure to either of the two run-out temperatures, ‘room temperature’ and ‘warmer climate’. It has been designed to assist clinicians make informed decisions about antibiotic prescribing in OPAT settings. It is also a starting point for further research specifically aimed at this setting.

Table 5

Conditions with verified stability data for refrigerated storage followed by a minimum run-out period of 24 hours at standard ‘room temperature’ or ‘warmer climate’ conditions


This systematic review demonstrates some drugs in current use in the home intravenous setting do not have chemical stability data for use in a warmer climate. Some drugs in current use have also been found to not have chemical stability to support use in OPAT under any conditions. These results should be considered alongside evidence for current patient outcomes within this setting. Further studies are needed to confirm the appropriate use of these medications in these settings and to ensure optimal possible patient outcomes.


Some proprietary data from manufacturing companies were not available for the review. As these data are not readily available to the rest of the scientific community the authors of this review could not verify these claims without reviewing the method of stability assessment.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
  29. 29.
  30. 30.
  31. 31.
  32. 32.
  33. 33.
  34. 34.
  35. 35.
  36. 36.
  37. 37.
  38. 38.
  39. 39.
  40. 40.
  41. 41.
  42. 42.
  43. 43.
  44. 44.
  45. 45.
  46. 46.
  47. 47.
  48. 48.
  49. 49.
  50. 50.
  51. 51.
  52. 52.
  53. 53.
  54. 54.
  55. 55.
  56. 56.
  57. 57.
  58. 58.
  59. 59.
  60. 60.
  61. 61.


  • EAHP Statement 3: Production and Compounding.

  • Contributors SJP, CL and NR performed the initial literature search and inclusion and exclusion criteria. All authors (SJP, CL, NR, TP and RF) then contributed to writing the paper.

  • Funding This research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent for publication Not required.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement All data relevant to the study are included in the article or uploaded as supplementary information.