Dispensing errors and ‘dispensing errors’

It just occurred to me that my posts on our recent paper discussing ‘dispensing errors’ could have been confusing for some clinicians or pharmacists for which ‘dispensing errors’ is totally altogether different..such as giving people the wrong drug which they have been prescribed.

Just thought I would clarify that up front. But things can get a bit fuzzy when pipettes are actually used to dispense/ give people drugs like in the example below. This goes to show the complexities of searching for a topic :

Drug Dev Ind Pharm. 2006 Feb;32(2):197-205.

Study of eye drops dispensing and dose variability by using plastic dropper tips.

Sklubalová Z, Zatloukal Z.


Department of Pharmaceutical Technology, Charles University in Prague, Faculty of Pharmacy, Hradec Králové, Czech Republic. zdenka.sklubalova@faf.cuni.cz


The application of eye drops from flexible dropper bottles fitted with different types of dropper tips is associated with the high variability of eye drop weights. The aim of this report was to investigate the simultaneous effect of three factors influencing the mean weight of drops dispensing from two plastic dropper tips. Using a designed experiment (Box-Behnken), the effect of the concentration of benzalkonium chloride solutions (BAC) in the range of 0-0.02%, the dispensing angle from 90 degrees to 30 degrees from horizontal, and the residual volume of liquid in the dropper bottle from 4 to 10 mL on the mean drop weights were examined. The significant effect of the increase in BAC concentration resulted in a linear decrease in drop weights for both of the dropper tips investigated. The significant effect of the dispensing angle was influenced by the dropper tip design. For the dropper tip A, the effect of the dropper tip tilt was described by the quadratic equation with a minimum, which corresponded to the dispensing angle equal to that of 48 degrees from horizontal. Below this angle, the increase in drop weights occurred due to the drop formation from the wetted external surface of the tip orifice. The linear decrease in drop weights in response to the decrease in dispensing angle was detected for the dropper tip B. The regression equations and the contour line plots obtained allowed the drop weights to be estimated for the actual combinations of both the BAC concentration and the dispensing angle. The effect of the residual volume was found to be non-significant. Based on the formula of Tate’s law, the direct proportion between surface tension of a solution and the radius of the effective perimeter of a dropper tip can be used to estimate the theoretical maximal weight of drops at the dispensing angle of 90 degrees . Using the stalagmometric values of surface tension of the BAC solutions, the maximal drop weights were estimated for both of the dropper tips investigated. A comparison between the theoretical and the experimentally measured drop weights enabled the dropper tips behavior to be discussed by using Harkins and Brown correction factor F. The F-value of 0.74 noted for the dropper tip A differed from that of stalagmometer F-value (0.61) indicating a deviation from the simple drop formation process in answer to more complicated design of the dropper tip A. On the other hand, the F-value of 0.6 observed for the dropper tip B demonstrated the better consistency with stalagmometry. As a result, the dropper tip B with the linear decrease of drop weights in response to the increased concentration of BAC and the decreased dispensing angle without the adverse external drop formation could be recommended in real drop dispensing.




Prompted by the interest in the paper and feedback from different people I have been doing a bit of digging to find other papers that might be of interest..Pipetting errors get a few mentions in PubMed abstracts, some very unusual topics – but there is not a lot of detail to infer whether the papers are relevant.


Food Addit Contam Part A Chem Anal Control Expo Risk Assess. 2012;29(11):1770-9. doi: 10.1080/19440049.2012.712063. Epub 2012 Aug 29.

Evaluation of variability and quality control procedures for a receptor-binding assay for paralytic shellfish poisoning toxins.

Ruberu SR, Langlois GW, Masuda M, Perera SK.


Drinking Water and Radiation Laboratory Branch and Environmental Management Branch, California Department of Public Health, 850 Marina Bay Parkway, Richmond, CA 94804, USA. Shiyamalie.Ruberu@cdph.ca.gov


The receptor-binding assay (RBA) method for determining saxatoxin (STX) and its numerous analogues, which cause paralytic shellfish poisoning (PSP) in humans, was evaluated in a single laboratory study. Each step of the assay preparation procedure including the performance of the multi-detector TopCount® instrument was evaluated for its contribution to method variability. The overall inherent RBA variability was determined to be 17%. Variability within the 12 detectors was observed; however, there was no reproducible pattern in detector performance. This observed variability among detectors could be attributed to other factors, such as pipetting errors. In an attempt to reduce the number of plates rejected due to excessive variability in the method’s quality control parameters, a statistical approach was evaluated using either Grubbs’ test or the Student’s t-test for rejecting outliers in the measurement of triplicate wells. This approach improved the ratio of accepted versus rejected plates, saving cost and time for rerunning the assay. However, the potential reduction in accuracy and the lack of improvement in precision suggests caution when using this approach. The current study has recommended an alternate quality control procedure for accepting or rejecting plates in place of the criteria currently used in the published assay, or the alternative of outlier testing. The recommended procedure involves the development of control charts to monitor the critical parameters identified in the published method (QC sample, EC₅₀, slope of calibration curve), with the addition of a fourth critical parameter which is the top value (100% binding) of the calibration curve.




Sex Transm Dis. 2012 Sep;39(9):733-4. doi: 10.1097/OLQ.0b013e318259efe2.

Mucus digestion improves the detection of Chlamydia trachomatis and Neisseria gonorrhoeae on the cobas 4800.

Miller A, Bromhead C, Jones M, Tustin P.


From Aotea Pathology Ltd, Wellington, New Zealand.


We compared the performance of cobas 4800 PCR for Neisseria gonorrhoeae with culture and found an association between pipetting errors and positive culture, caused by mucopurulent discharge. The problem may be resolved by pretreating samples with Sputasol, thus improving detection of both Chlamydia trachomatis and Neisseria gonorrhoeae.




Rev Sci Instrum. 2010 Dec;81(12):124301. doi: 10.1063/1.3509395.

Absorbance and fluorometric sensing with capillary wells microplates.

Tan HY, Cheong BH, Neild A, Liew OW, Ng TW.


Laboratory for Optics, Acoustics, and Mechanics, Department of Mechanical and Aerospace Engineering, Monash University, Clayton, Victoria 3800, Australia.


Detection and readout from small volume assays in microplates are a challenge. The capillary wells microplate approach [Ng et al., Appl. Phys. Lett. 93, 174105 (2008)] offers strong advantages in small liquid volume management. An adapted design is described and shown here to be able to detect, in a nonimaging manner, fluorescence and absorbance assays minus the error often associated with meniscus forming at the air-liquid interface. The presence of bubbles in liquid samples residing in microplate wells can cause inaccuracies. Pipetting errors, if not adequately managed, can result in misleading data and wrong interpretations of assay results; particularly in the context of high throughput screening. We show that the adapted design is also able to detect for bubbles and pipetting errors during actual assay runs to ensure accuracy in screening.




J Biomol Screen. 2010 Oct;15(9):1160-4. doi: 10.1177/1087057110381648. Epub 2010 Sep 17.

Capillary wells microplate with side optical access.

Tan HY, Ng TW, Neild A, Liew OW.


Laboratory for Optics, Acoustics, & Mechanics, Department of Mechanical & Aerospace Engineering, Monash University, Clayton, Victoria, Australia.


The presence of bubbles in liquid samples residing in microplate wells causes inaccuracies in fluorescence measurements. In addition, pipetting errors, if not adequately managed, can result in misleading data and wrong interpretations of assay results, particularly in the context of high-throughput screening. In this work, the authors describe an adapted design to the capillary wells microplate approach that permits side viewing. They demonstrate a prototype that detects bubbles and pipetting errors during actual assay runs to ensure accuracy in screening.




I am also keen to see this recent paper :

J Lab Autom. 2013 Jan 25. [Epub ahead of print]

Evaluation of a Liquid Dispenser for Assay Development and Enzymology in 1536-Well Format.

Butendeich H, Pierret NM, Numao S.


1Novartis Institutes for Biomedical Research, Basel, Switzerland.


Although developments in liquid dispensers have made the use of 1536-well plates for high-throughput screening (HTS) standard, there is still a gap in dispenser technology for performing matrix experiments with several components. Experiments such as those performed during assay development and enzymological studies are therefore still performed by manual pipetting in lower-density plates. We have evaluated a new dispenser, the Certus liquid dispenser (Gyger Fluidics GmbH, Switzerland), that is capable of flexible dispensing in 1536-well format, with a dead volume of less than 200 µL. Taking advantage of the precision of the dispenser for volumes down to 50 nL, we have created concentration gradients on plates by dispensing different volumes of reagent and then backfilling with buffer. Using this method and the flexibility of the dispenser software, we have performed several multidimensional experiments varying two to three components, including an assay development for an HTS, a mode of inhibition study, and a cofactor optimization, in which we determined 32 K(M) values. Overall, the flexibility of the plate layout for multiple components, the accuracy to dispense volumes ranging 2 log orders, and minimal reagent usage enable this dispenser for complex biochemical experiments.




..And also his one in which a 300,000 compound screen resulted in no hits. It makes me wonder whether the dispensing approach could have an impact on screens like this?


J Ocul Pharmacol Ther. 2013 Apr;29(3):298-309. doi: 10.1089/jop.2012.0135. Epub 2012 Oct 25.

In pursuit of synthetic modulators for the orphan retina-specific nuclear receptor NR2E3.

Qin Q, Knapinska A, Dobri N, Madoux F, Chase P, Hodder P, Petrukhin K.


Department of Ophthalmology, Columbia University Medical Center, New York, NY 10032, USA.



NR2E3 is an orphan nuclear receptor expressed exclusively in photoreceptor cells of the retina. NR2E3-specific modulators may prolong photoreceptor survival in patients with dry age-related macular degeneration and other forms of retinal degeneration. To definitively establish NR2E3 as a photoreceptor protection target, identification of small-molecule NR2E3 modulators and their testing in animal models of retinal degeneration are required. Development of the high-throughput screen (HTS)-compatible screen for small-molecule NR2E3 modulators is the first step toward this goal.


Purification protocol for isolation of the functionally competent soluble NR2E3 protein after its expression in the insect Sf9 cells was developed. The time-resolved fluorescence energy-transfer (TR-FRET) assay assessing agonist-sensitive interaction between apo-NR2E3 and transcriptional corepressor RetCOR was used for characterization of the previously reported putative NR2E3 agonist, Compound 11a, and to conduct the HTS for novel small-molecule NR2E3 modulators (direct and inverse agonists). A counterscreen TR-FRET assay that measures the affect of test compounds on PPARγ interaction with corepressor NCOR was used for assessing the specificity of compounds identified in the HTS.


We developed the cell-free TR-FRET assay for small-molecule NR2E3 modulators, which is based on agonist-induced disruption of the interaction between GST-tagged apo-NR2E3 and MBP-tagged fragment of transcriptional corepressor RetCOR. Compound 11a, a putative NR2E3 agonist, did not affect the NR2E3-RetCOR interaction, as was established by its titration in the developed assay. The assay was miniaturized for an ultralow-volume 1,536-well format and automated into 3 simple pipetting steps. Consistent with excellent assay performance, the test runs established a Z’-score within the 0.6-0.8 range. Analysis of the mid-size National Institutes of Health collection of 315,001 structurally diverse drug-like compounds confirmed excellent assay performance, but did not reveal NR2E3-specific agonists or inverse agonists.


A robust and reliable TR-FRET assay for small-molecule NR2E3-specific modulators suitable for the analysis of million compound-strong HTS libraries was developed. A previously described putative NR2E3 agonist, Compound 11a, is unlikely to represent a direct NR2E3 agonist. Application of the developed assay for screening of a more abundant and diverse compound collection be required for identification of synthetic NR2E3 ligands.



[PubMed – in process]



Free PMC Article


Well if anyone reading this has any papers on dispensing errors that I might be interested in – as opposed to those documenting “giving the wrong doses of drugs to patients” – please let me know and add them to the comments below. My guess is I will be digging for a good long while at this rate.


No comment yet

1 ping

  1. Joe says:

    The problem with “pipetting errors” is that it is a broad brush that covers much but means little. Everybody will accept that a pipette volume can be off by ~10%. So they all figure that multiple errors will lead to a possible error of 50-70%. For many biological applications (e.g., IC50) people feel that this is OK. But that broad brush has just obscured the much more important problem elucidated in your PLOS ONE article–liquid transfer errors can generate errors that are orders of magnitude in scale. In the PLOS ONE paper is an example of a difference in IC50 of 250-fold. That is not 250% (still a very large number that should cause discomfort) but 25000%. But people ignore it because it is “just a pipetting error.” I see this a bit like someone searching for his keys under a street lamp even though he lost the keys 50 meters away. The light is better. If the only tool you have is a pipette (automated or manual) and you are following your labs SOPs, then the answer you get is the best possible answer. We have seen people complain about the cost of acoustic dispensers as they spend hundreds of thousands pounds Sterling on reagents, manpower and tips to get answers that have little or no utility. But they can point to large amounts of data accumulated and check off items for their annual reviews.

Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>