Benzyl Alcohol at 0.9%: Where That Figure Comes From and Why the 28-Day Clock Exists

Both figures come from outside the vial. The 0.9% is a manufacturer's declared formulation attribute printed on the label — 9 mg/mL of benzyl alcohol — and the same compendial article is also distributed at 1.1% (11 mg/mL) under the identical product name, which by itself proves 0.9% is not a pharmacopoeial requirement ^[1]. The 28 days is the last enumeration point of the compendial preservative test: antimicrobial effectiveness testing runs to day 28 and stops, so day 28 is the last day on which any published evidence about that preservative system exists ^[2][3]. Practice standards then adopted the end of the test window as the default ceiling on an opened container ^[6][7].

That answer is more useful than the folklore version, because it tells you what each number does and does not cover. This article treats preserved diluent as laboratory material: a compendially defined vehicle used at the bench with lyophilised research compounds, which are supplied for research use only and are not medicinal products.

Where the 0.9% figure actually comes from

Start with the label, because the label is the specification. Bacteriostatic Water for Injection, USP is described as a sterile, nonpyrogenic preparation of water for injection containing 0.9% (9 mg/mL) or 1.1% (11 mg/mL) of benzyl alcohol added as a bacteriostatic preservative, with a pH of 5.7 across a range of 4.5 to 7.0, supplied in multiple-dose containers from which repeated withdrawals may be made ^[1]. Two concentrations, one monograph, one product name. Read the carton.

Benzyl alcohol is one of the small set of preservatives that dominates licensed parenteral products, chosen on documented precedent rather than novelty ^[11]. The concentrations actually cleared for a given route are recorded in FDA's Inactive Ingredient Database, the checkable source when someone asserts a figure is standard ^[12] — and what that record shows is a band, not a constant.

The reason a formulator sits at the low end of that band is regulatory and explicit. ICH Q6A requires that the lowest specified concentration of antimicrobial preservative be demonstrated to be effective using a pharmacopoeial effectiveness test, that effectiveness be shown during development, during scale-up and throughout shelf life, and that acceptance criteria for preservative content be based on the levels needed to maintain microbiological quality at every stage of the product's proposed usage; content is then normally a release test in its own right ^[5]. So the number is the output of a constrained optimisation: as little preservative as will still pass, because a preservative is a biologically active excipient carrying its own toxicity and compatibility costs, not a free safety margin.

What "bacteriostatic" means at that concentration

The word is precise and it is not a synonym for sterile. A bacteriostatic concentration suppresses the multiplication of vegetative organisms; it is not sporicidal, it does not sterilise a compromised container, it does nothing about the endotoxin a bacterial population leaves behind, and it does not act on the timescale of a single closure entry — the compendial criteria demand reductions measured in days, not minutes ^[2]. Efficacy is also concentration-dependent rather than binary, which is why preservative content is written as a specification range and why anything that lowers the concentration actually in solution is a quality question rather than a rounding error ^[5][11]. A formulation that passes at its lowest specified concentration has, by construction, very little headroom below that point.

The test that decides whether 0.9% works: USP <51>

Antimicrobial effectiveness testing is a challenge test, and its design is worth knowing because every claim about a preserved diluent traces back to it. Five organisms are used: Staphylococcus aureus ATCC 6538, Pseudomonas aeruginosa ATCC 9027, Escherichia coli ATCC 8739, Candida albicans ATCC 10231 and Aspergillus brasiliensis ATCC 16404. Each is grown to a standardised suspension of roughly 10⁸ colony forming units per mL and inoculated separately into the product to give a challenge of 10⁵ to 10⁶ CFU/mL; the containers are then held at 20–25 °C and survivors enumerated at 7, 14 and 28 days ^[3].

For USP Category 1 products — injections and other parenterals — the criteria for bacteria are a reduction of not less than 1.0 log from the initial calculated count at 7 days, not less than 3.0 log at 14 days, and no increase from the 14-day count at 28 days. For yeasts and moulds the requirement is no increase from the initial calculated count at 7, 14 and 28 days, where "no increase" means not more than 0.5 log above the previous measured value ^[2]. Those are floors: passing tells you a formulation cleared a minimum, not how much margin it cleared it by.

Why Europe's version of the same test is harder

The three major pharmacopoeias use the same organisms and essentially the same inoculum preparation and execution; what they do not share is acceptance criteria ^[3]. The European Pharmacopoeia samples earlier and more often — 6 hours, 24 hours, 7, 14 and 28 days, against the USP's three points ^[3]. For parenteral preparations, Ph. Eur. criterion A requires a 2 log reduction in bacteria at 6 hours, 3 log at 24 hours and no recovery at 28 days, with fungi reduced by 2 log at 7 days and no increase thereafter. Criterion B — 1 log for bacteria at 24 hours, 3 log at 7 days, 1 log for fungi at 14 days — applies only where A cannot be attained for justified reasons, such as increased risk of adverse reactions from a more aggressive preservative system ^[4].

So "passes preservative effectiveness testing" is an incomplete statement. A formulation can satisfy USP Category 1 and fail Ph. Eur. criterion A, because the European test asks for meaningful kill within the first day and the American one measures nothing before day 7 ^[2][3][4]. Where the specification matters to your documentation, name the chapter and the criterion.

Where the 28-day clock comes from

Line the numbers up and the origin is obvious: both compendial tests end at 28 days ^[2][3]. USP General Chapter <797> then holds that a conventionally manufactured multiple-dose container that has been entered or punctured is not used beyond 28 days unless the manufacturer's labelling specifies a different period ^[6]. CDC injection-safety guidance says the same in operational language — date the container at first puncture, discard within 28 days unless the manufacturer states another date, and never carry the in-use date past the printed expiry ^[7].

Being plain about what is not published: neither the compounding chapter nor the CDC guidance offers a derivation of 28 days from puncture-frequency or contamination data for any particular product. It is a default ceiling inherited from the length of the test window, applied where product-specific in-use data does not exist — a defensible convention, not a measurement. Treating it as a measured property of the fluid is the most common error in this area.

The alternative route to the same kind of number is experimental. European in-use stability guidance establishes the period for a multi-dose product from study on the finished product itself, with containers subjected to the conditions of normal use and testing run over at least the proposed in-use period, so the figure is derived rather than assumed ^[9]. Where a manufacturer has done that work the label period supersedes the default — which is precisely the escape clause both USP <797> and CDC write into their own wording ^[6][7].

Why the container is 30 mL rather than a bottle

The volume cap is the third control on the same risk, and FDA states the rationale directly: limiting the volume of a multiple-dose vial limits the number of septum punctures, which limits the opportunity to compromise container integrity and introduce contamination ^[8]. Preservative concentration, container volume and in-use period are one system with three levers; weakening any one transfers load to the other two.

The empirical contamination rate in real multi-entry containers is low but not zero. A prevalence study collected every in-use multiple-dose vial in a 1,300-bed hospital on a single day and found one vial and one spike contaminated with Staphylococcus epidermidis out of 227 — 0.9%, 95% confidence interval 0.3 to 2.1%. The study was prompted by two deaths from Pseudomonas aeruginosa meningitis traced to a contrast medium used as a multiple-dose vial over eight days ^[10]. That is hospital epidemiology, not a bench simulation, and it is the clearest illustration of why multi-entry containers are governed by hard limits rather than judgement.

What can put a container below its effective concentration

• Sorption. Preservatives interact with other formulation components and can partition into elastomeric closures across shelf life, lowering the concentration actually in solution below the labelled figure ^[11].
• Headspace. Benzyl alcohol is volatile enough to distribute into an expanding headspace as liquid is withdrawn — a real term in the mass balance of a partly used container, and one the sealed-container test does not create ^[2][11].
• Temperature. The compendial test runs at 20–25 °C ^[3]; a container held colder or warmer sits outside the condition under which its preservative system was demonstrated.
• pH. The specification runs from 4.5 to 7.0 ^[1], and preservative activity and compatibility both vary across that span, as do the competing pH dependencies that govern peptide stability, which rarely share an optimum. A reconstituted solution sits at whatever pH the combined system produces, not at the diluent's.
• Reconstitution itself. Once a compound is dissolved in preserved diluent, the result is a new formulation with its own preservative concentration and its own microbiological behaviour, and no effectiveness test has been run on it.

That last point is where the published record quietly stops, and it should be stated rather than glossed. Compendial preservative data covers the diluent as supplied; there is no compendial in-use data for research peptides reconstituted in preserved diluent, because no one has run the test on those combinations. Nor does the figure speak to the separate, chemical question of how long a peptide holds up once it is in solution, which turns on temperature, pH and sequence rather than on microbiology. Presenting the 28-day figure as an assurance about a reconstituted research solution extends a result well past the system it was measured on.

Recording practice that makes the numbers defensible

• Record the preservative concentration from the carton rather than assuming it. 0.9% and 1.1% material carry the same product name and are different formulations ^[1].
• Write the puncture date on the container at first entry, and derive the in-use limit from that date, capped by the labelled period where one exists and by the printed expiry in every case ^[6][7].
• Keep the diluent manufacturer, lot and expiry in the same record as the compound lot, and log storage temperature for in-use containers — the effectiveness data underpinning the limit was generated at 20–25 °C ^[3].
• Use a fresh needle for every entry and swab the closure with 70% isopropyl alcohol, allowing it to dry; the preservative works on a scale of days and cannot compensate for a poor entry ^[2][7].
• Discard on any deviation — cloudiness, discolouration, particulate, an undated container, a suspected integrity failure — regardless of days remaining ^[7].
• Name the compendium and the criterion when recording a preservative specification, because USP Category 1 and Ph. Eur. criterion A are different bars ^[2][4].

Preserved diluent is useful in a laboratory precisely because its attributes are stated, testable and traceable to a published chapter rather than to a vendor's description — and knowing which chapter, and where its evidence stops, is what turns a number on a carton into a record you can defend. The materials discussed here, and the research compounds such diluents are used with at the bench, are supplied for research use only. They are not medicinal products, they are not intended for administration to humans or animals, and nothing above should be read as guidance for anything other than handling laboratory material.