A set of vials can have the expected average content while individual vials differ appreciably. Between-vial testing asks how those unit values vary. To interpret the answer, the report needs to preserve individual results and explain how much variation could come from the measurement itself.
Start with the quantity being compared
Content per vial, peptide mass fraction and solution concentration are different quantities. A set of vials could have similar composition by mass but different amounts of material. Conversely, similar total powder masses would not establish equal amounts of the target peptide.
The National Research Council of Canada’s ANGII-1 certificate describes a homogeneity assessment across representative units from its filling series. Its certified quantity is angiotensin II mass fraction; this is a real example of assessing unit differences, not evidence of total peptide content in Novum vials.NRC — ANGII-1 certified reference material certificate (opens in a new tab)
| Reported quantity | What must stay consistent |
|---|---|
| mg per vial | Definition of the target and recovery from the original vial |
| mg per g of material | Mass-fraction basis and material conditioning |
| mg per mL | Preparation volume and which solution was measured |
Do not convert one type of homogeneity claim into another without the additional measurements that connect them.
Look beyond the average
Consider two original illustrative sets of three content results. Set A contains 9.8, 10.0 and 10.2 mg per vial. Set B contains 6.0, 10.0 and 14.0 mg per vial. Both average 10.0 mg; their observed ranges are 0.4 and 8.0 mg respectively.
The mean communicates the centre of the selected results. The range communicates the difference between their largest and smallest values. Neither statistic, on its own, establishes the distribution of every untested vial.
A standard deviation or relative standard deviation can summarise spread, but it should accompany the sample size and the measurement basis. A precise-looking percentage calculated from very few units is still based on very few units.
Keep any acceptance criteria separate from this description. The arithmetic demonstrates different spreads; it does not supply a universal acceptable content range.
Separate material variation from measurement variation
Different measured values can reflect differences in the vials, differences introduced during preparation, instrument variation, or a combination. The experiment needs enough structure to examine these contributions rather than assigning all observed spread to filling.
Repeated injections from one prepared solution mainly describe that solution’s analytical repetition. Independently prepared samples introduce another level of variation. Separately selected original vials provide the unit comparisons needed for the between-vial question.
In the ANGII-1 work, the certificate describes an analysis of variance and a homogeneity contribution to uncertainty. This shows how a reference-material assessment can acknowledge unit variation within the reported uncertainty rather than declaring it absent.NRC — ANGII-1 certified reference material certificate (opens in a new tab)
If every vial is prepared only once, preparation error and genuine vial differences may be difficult to separate. A sophisticated statistical label cannot recover experimental comparisons that were never made.
Keep the conclusion proportional to the design
Ask which vials were selected, whether they span the relevant filling sequence, and whether any results were excluded. A tightly grouped set from one small part of a run does not automatically describe all its other parts.
Report the observed result first: for example, that the selected units had a stated mean and spread under a specified method. Any wider batch inference should identify the sampling rationale and the assumptions that support it.
If a follow-up result differs, compare methods and preparation before calling it a manufacturing change. The practical objective is to locate the source of variation, not simply to choose the more reassuring summary.
Sources and further detail
- NRC — ANGII-1 certified reference material certificate (opens in a new tab)
Primary certificate, homogeneity and uncertainty sections read. Its mass-fraction study is not a per-vial fill study. Numerical vial comparisons here are original illustrations.
Sources checked 19 September 2026. Worked examples are illustrative unless a supplied report is explicitly identified. This article has not undergone independent scientific peer review.