POST 00316 : ESTIMATING WASTAGE FACTORS

Post00316 ESTIMATING WASTAGE FACTORS 29 January 2001 CONTENTS 1. ESTIMATING VACCINE WASTAGE FACTORS Continued In Technet Post00311, ESTIMATING VACCINE WASTAGE FACTORS, 15 January 2001, a discussion on estimating vaccine wastage and vaccine wastage factors was posted. The context of the discussion was the need for GAVI, the Global Alliance on Vaccines and Immunization, to look at vaccine wastage - with its potential for enormous unsustainable vaccine costs. The other discussion context was and remains the need for national programs to monitor and appropriately manage vaccine wastage to contain costs and enhance program sustainability. In today's posting, Robert Steinglass, BASICS, provides some clarification and suggests that the order of the messages posted was misleading. Robert's final paragraph was not copied to Technet and was not posted - it does help a lot. Julie Milstien, WHO/ATT, Souleymane Kon鬠WHO/AFRO, and Bob Davis, UNICEF, Anton Luchitsky, PATH, and Hans Everts, WHO/EPI, add to the discussion. Opinion, comments and additions please: [[email protected]][email protected][/email] or use your reply button ___________________________________________________________________________ Moderators Note Apologies to Technet Forum subscribers if it was difficult to make sense of the discussion(s) in Post00311. The messages were grouped thematically, as some messages were on estimating wastage factors while others were related to methods of estimating wastage and the feasibility of programs routinely estimating wastage. It seemed to make more sense thematically rather than chronologically. The final versions of the formula for the calculation of wastage factors as modified by Robert Steinglass was posted as the lead item in the discussion in Post00311 and is repeated below. Our thanks to Michel Zaffran, WHO/V&B, who kindly translated the message from Souleymane Koné¡Š___________________________________________________________________________ Date: Fri, 19 Jan 2001 From: "Robert Steinglass" To: [[email protected]][email protected][/email] Subject: calculation of wastage FACTORS (not rates) Dear Allan, I think an important point may well be overlooked, but first a quick divergence to correct the record. The order in which the postings appeared in the last TECHNET Forum is misleading. The wastage formula which is posted at the start is what emerged after I provided edits on an earlier version of John Lloyd's formula. Therefore, my critique which comes immediately after the formula does not make a lot of sense, since I am critiquing a different formula than the one which appears. The real point which is worth stressing to the TECHNET readers and should not be overlooked is that the proposed wastage formula would be used to directly calculate wastage FACTORS, thereby bypassing calculation of wastage RATES. * Missing from the thread of e-mails was my following comment to John Lloyd, which may shed light for TECHNET Forum readers on the new proposed formula: "The alternative to estimating vaccine wastage which I like best is what they do in the NIS. They directly derive a wastage factor (which is much easier to do arithmetically and which is the key coefficient which is really needed to forecast vaccine requirements). They don't bother with complicated calculations of wastage rates which would have to be converted in any event into wastage factors. So they can tell you from their stock records that, e.g., 1.5 doses of DPT were required to administer a single dose. Pretty simple." Robert --- From: [[email protected]][email protected][/email] Date: Tue, 16 Jan 2001 Subject: RE: Post00311 ESTIMATING VACCINE WASTAGE FACTORS To: Technet Moderator I have to agree with Mark. Trying to develop hard and fast rules for calculations for all situations is a fruitless expenditure of time. No one understands the use of a formula better than someone who has figured it out for him or herself. However, it would be useful for GAVI to standardize what they want reported and the considerations that would determine how it was calculated (stock levels, vials lost because of expiry or cold chaing break, etc). I am confused in this discussion because although the dates look like the order is chronological, it also looks like John's first message is his final word or else I cannot find his final word. What are you trying to say, bottom line, John? Dr Julie Milstien Coordinator, Access to Technologies Department of Vaccines and Biologicals World Health Organization 20 Ave Appia Room M231 1211 Geneva 27 Switzerland tel 41 22 791 3564 fax 41 22 791 4384 email [[email protected]][email protected][/email] --- Date: Tue, 16 Jan 2001 From: [[email protected]][email protected][/email] (Robert Davis) Subject: Re:Post00311 ESTIMATING VACCINE WASTAGE FACTORS To: Technet Moderator , Dear All, I like a very simple method: Usage is doses administered/doses used or discarded. Wastage is 100 % minus usage. This is very simple to calculate, requiring only one division and one subtraction, well within the intelligence levels of agency directors, for example. Bob Davis UNICEF/Nairobi --- From: Souleymane Koné ¦lt;[email protected]> To: Subject: Re: Post00311 ESTIMATING VACCINE WASTAGE FACTORS Date: Fri, 19 Jan 2001 Dear Allan, Would you pleased translate my contribution to this important issue. [The English version of the text follows. Our thanks to Michel Zaffran, WHO/V&B, who kindly translated this message over the weekend!] Le present sujet sur l'estimation des facteurs de perte des vaccins est d'une importance capitale. De facon general, deux indicateurs mesurent les pertes de vaccins: le taux de pertes et le facteur de pertes. Les deux indicateurs sont en fait lié³ et peuvent etre obtenus l'un a partir de l'autre. Ce qui en mon sens les differencie c'est l'utilisation pratique directe de l'un ou de l'autre dans la gestion des vaccins. Le facteur de perte sera plus utilise dans la planification pour le calcul des besoins en vaccins. En d'autres termes de combien de doses de vaccins aurai-je besoin pour chaque cible a vacciner. Le taux de perte indiquera plus facilement l'efficacite dans l'utilisation des vaccins. En d'autres termes quelle proportion des doses de vaccins destines a notre cible ne l'atteignent pas effectivement, c'est-a-dire perdue. La gestion du programme devrait bien entendu rechercher les sources et les causes des pertes afin de les minimiser. Pour l'estimation des indicateurs de perte des vaccins (taux ou facteur de pertes) ci-dessus mentionnes, l'essentiel reside dans l'esprit, dont les formules ne sont (devront etre) l'illustration. Ainsi la "premiere formule" de John du facteur de pertes est correcte: Fw = ((B+C) x D)/A, le taux de pertes sera donc: Rw = 1 - A / ((B+C) x D) Cependant deux clarifications sont necessaires: 1. le numerateur ((B+C) x D) devra tenir compte des differentes tailles de flacons utilisees sur le terrain (observations faites par R. Steinglass). Certains pays utilisent pour certains antigenes differentes tailles de flacons (10 doses et 20 doses). Dans pareils cas les informations sur le nombre de flacons correspondant a chaque taille devra etre disponible pour les calculs. Ainsi le numerateur sera: {(B1+C1) x D1 + (B2+C2) x D2) + ...} ou: B1 = nombre de flacons de taille 1 ouverts pour utilisation C1 = nombre de flacons de taille 1 abimes pour D1 = nombre de doses par flacon de taille 1 B2 = nombre de flacons de taille 2 ouverts pour utilisation C2 = nombre de flacons de taille 2 abimes pour PCV, ch. froid D2 = nombre de doses par flacon de taille 2 2. le denominateur: A (nombre de doses administrees) devra etre relatif a la cible vaccinee. Si c'est le cas dans certains pays la vaccination des hors cibles atteint parfois selon les antigenes 10 voire 15% du total vaccine, par consequent les doses administrees aussi. D'ou une sous-estimation du facteur de perte des vaccins si on utilisait au numerateur le nombre de doses administrees (total population vaccinee) au lieu du nombre de doses administrees a la cible (total cible vaccinee). La "deuxieme formule" de John: Fw = ((B+C-D) x E)/A, est aussi correcte, et le taux de pertes sera de: Rw = 1 - A/((B+C-D) x E) Cependant le contenu du terme "C" (nombre de flacons sorties du depot for utilisation) devra etre plutot "nombre de flacons recus". Le numerateur ((B+C-D) x E) designera la quantite de vaccins mobilises pour la vaccination. Les memes observations faites pour la premiere formule s'appliquent pour la deuxieme. Meilleures salutations Souleymane ___________________________________________________________________________ English version of Souleymane Kon駳 message. Dear Allan, The issue of how to estimate vaccine wastage rates is a most important one. Generally speaking, two numbers can be used to measure vaccine wastage: the wastage rate and the wastage factor. The two numbers are related and one can be calculated from the other. In my opinion what differentiates the two numbers is how they are actually used in practice to manage vaccine stocks. The wastage factor is normally used to plan and calculate the quantities of vaccines needed. In other words: "how many doses of vaccines will I need for a given targeted population?" The wastage rate indicates more directly the efficiency of vaccine use. In other words, which proportion of vaccine doses do not get used for the target population and is wasted. The programme management must of course identify the reasons and causes for this wastage in order to minimise them In order to calculate the two numbers, the most important is their meaning rather than the formula: John?s first formula for the wastage factor is correct: Fw = ((B+C) x D)/A, The wastage rate will therefore be: Rw = 1 - A / ((B+C) x D) However two clarifications are necessary_ 1. The numerator: ((B+C) x D) must take account of the different vial sizes used in the filed (this was pointed out by Robert Steinglass). Indeed, some countries use, for certain antigens, different vial sizes (10 and 20 doses). In these cases, the number of vials of each type of vial size should be available for the calculation. The numerator will therefore be: {(B1+C1) x D1 + (B2+C2) x D2) + ...} Where B1 = number of vials of size 1 opened for use C1 = number of vials of size 1 with VVM expired D1 = number of doses in vials of size 1 B2 = number of vials of size 2 opened for use C2 = number of vials of size 2 with VVM expired D2 = number of doses in vials of size 2 2. The denominator: A (number of doses administered) should be adjusted to the number of doses actually administered to the target population. In some countries and for some antigens, the group vaccinated includes sometimes 10 to 15% of children, which are not in the target group. Therefore the number of doses actually administered may be 10 to 15% higher than those really administered to the target group. If we were to use the number of doses administered, without adjusting for target population, the wastage rate would be under-estimated John?s second formula: Wastage factor Fw = ((B+C-D) x E)/A, is also correct and the wastage rate will be Rw = 1 - A/((B+C-D) x E) However, the meaning of "C" (number of vials taken out of stock for use) should instead be "number of vials received" the numerator, ((B+C-D) x E) then will refer to the total quantity of vaccine used for the vaccination session Best regards Souleymane Kone --- From: "Anton Luchitsky" To: "Technet Moderator" Subject: Re: Post00311 ESTIMATING VACCINE WASTAGE FACTORS Date: Tue, 23 Jan 2001 Dear colleagues, Agreeing with Mark Weeks's observations (particularly about the importance of keeping indicators simple) I would like to mention that in Ukraine use of a very simple indicator, namely vaccine wastage coefficient (total doses used divided by the total number of vaccinations given during a specified period of time) has proved to be very valuable; we demonstrated that over the past 1-2 years wastage of vaccines decreased considerably as a result of changes in immunization practices (grouping of children for immunization session, arranging outreach sessions, etc.) as well as more rational distribution of vaccines and prompt investigation of causes of high wastage. Both the numerator (doses used) and denominator (vaccinations given) are routinely recorded in monthly reports prepared by health settings involved in immunization and subsequently analyzed at immunization points, in rayons and at higher levels (oblast, MoH). When developing this indicator the local working groups of experts decided that the numerator should include the sum of doses utilized during immunization and doses written off (e.g., frozen or expired), the latter at the time believed to be insignificant and quite rare. It is possible that the working group may decide to recommend separate recording and routine analysis of the proportion of written off vaccine (in unopened vials). Vaccine wastage coefficient can work very well if the country (or oblasts) predominantly use only one presentation of a given vaccine, which is often typical in cases of centralized procurement of biologicals. The accuracy of this indicator may be to some extent compromised when vaccine in various presentations (e.g. in 2- and 10-doses vials) is used on the same territory; besides, it is difficult to "compare" wastage if one rayon is using 2-dose vials and another one 10-dose vials..... Whereas it is technically possible to account for different vial sizes when calculating wastage(s), the added complexity may not be always justified - immunization managers are mainly concerned about the price of one vaccination and not about the wastage itself. Here is a possible way out for the situation like this 1. Continue monitoring "average" wastage coefficient for given rayons/oblasts as described above. 2. Determine the "average" price of one dose of vaccines used on the given territories during the given period. [for example oblast X used 40,000 doses DPT during a 6-month period: 30,000 doses ($0.12 each) in 2-dose vials and 10,000 doses ($0.08 each) in 10-dose vials). The average price of one dose was (30,000 x 0.12 + 10,000 x 0.08)/ 40,000 = $0.11 3. Multiplying "average" wastage coefficient by "average" price of one dose determine the cost of one immunization given on this territory. 4. Analyze the correlation between average price of one dose and average wastage on the given territory. Determine the optimal proportion of vial sizes for this territory [increase of the average cost of a dose due to increased proportion of smaller size vials is only justified as long as this results in decrease of the average wastage coefficient by the same degree or more]. Best wishes, Anton Luchitsky, M.D. Program Officer Program for Appropriate Technology in Health (PATH) Khreshchatyk St., 29, Apt. 50 Kiev, 01001, Ukraine Phone: (380)-44-462-0366 FAX: (380)-44-462-0365 E-mail: [[email protected]][email protected][/email] Visit the PATH Web site at http://www.path.org --- To: [[email protected]][email protected][/email] Subject: RE: Post00311 ESTIMATING VACCINE WASTAGE FACTORS Date: Thu, 25 Jan 2001 Dear Allan, I just read the discussion in Technet Forum 00311. I hope my comments do not come too late or are dealt with in the attachments mentioned in the discussion. Whenever it says 'Doses administered' it should say 'Doses administered according to EPI schedule'. A dose given to the child at the wrong age group is a wasted dose. The C in John's second formula should not be the number of doses issued from the store, because that would lead to a double count, since they are included in the number of vials in stock. The C should be the number of vials received over that period. The formula would then be: ((B + C - D) x E) - A)/ A x 100, A = Number of children immunized according to EPI schedule over period Y B = Number of usable vials in stock at the beginning of period Y C = Number of vials received during period Y D = Number of usable vials in stock at the end of period Y E = Number of doses per vial An inconvenience of calculating on basis of vials rather than doses is that remaining doses in opened vials will be counted as lost even though they may be used the day after the inventory was made. In smaller health centres, where few vials are used, that may have an important impact on the wastage rate. Regards Hans Everts Technical Officer EPI WHO Geneva Tel: 00 41 22 791 3683 ___________________________________________________________________________ ESTIMATING VACCINE WASTAGE FACTORS Robert Steinglass' final version 1. Using National (or Sub-national) Records of Doses Administered Existing language would need to be modified Formula is: Wastage factor = ((B + C) x D)/ A Where : A = the number of doses administered B = the number of vials opened for use C = the number of closed vials discarded due to cold chain failure or VVM indication D = the number of doses per vial 2. Using Stock Records Again modify the existing language. Formula would be : Wastage factor = ((B + C - D) x E)/A Where : A = the number of doses administered B = the number of usable* vials in stock at the start of the year C = the number of vials issued from the store for use D = the number of usable* vials in stock at the end of the year E = the number of doses per vial This could then be followed by a simple conversion from wastage factor to wastage rates, since people are still accustomed to report their wastage rates. * Vials of vaccine which are still permitted to be used. ____________________________________*________________________