Drug shortages : 39% of French people at risk in 2024

39% of French people say they expe­rien­ced medi­cines shor­tages in 2024¹ – up from 44% in 2023 – and 35% of them say they were unable to obtain alter­na­tive treat­ments. To quan­ti­fy the extent of the shor­tage in France, the DREES (Direc­tion de la recherche, des études, de l’évaluation et des sta­tis­tique) conduc­ted an ori­gi­nal stu­dy at the request of the ANSM (Agence natio­nale de sécu­ri­té du médi­ca­ment et des pro­duits de san­té) using data from labo­ra­to­ries and wholesalers/distributors. Clé­ment Dher­bé­court, an expert in public poli­cy eva­lua­tion at the DREES and co-author of the stu­dy, details the main findings.

Clé­ment Dher­bé­court. We had two main objec­tives : the first was to improve the accu­ra­cy of the ten­sion signal on drug pro­duc­tion by pro­du­cing dai­ly ana­lyses of the num­ber of sto­ckouts and sto­ckout risks within labo­ra­to­ries, as well as the sales volumes to phar­ma­cies affec­ted by these ten­sions. The second objec­tive was to esti­mate the change in the num­ber of “mis­sing boxes”, i.e. boxes that could not be sold to phar­ma­cies by their sup­pliers, which is cur­rent­ly the most rele­vant indi­ca­tor of the shor­tage situation.

To do this, we relied on two main data sets. DREES had figures for sales of medi­cines to phar­ma­cies, either by who­le­sa­lers-dis­tri­bu­tors (who account for 80% of the volume of boxes sold) or direct­ly by labo­ra­to­ries. ANSM gave us access to the reports of shor­tages and risks of shor­tages sub­mit­ted by laboratories.

Not at present. To get a bet­ter idea of this figure, we would need to rely on an elec­tro­nic pres­crip­tion sys­tem, which is only just begin­ning to be deve­lo­ped. Alter­na­ti­ve­ly, we could consi­der more indi­rect data, such as figures on orders pla­ced by phar­ma­cies for indi­vi­dual medi­cines. This data exists but is not cur­rent­ly publi­cly avai­lable. The 2025 Social Secu­ri­ty Finan­cing Act², which has been pas­sed but has not yet come into force, creates a public sys­tem for acces­sing phar­ma­cy order data to enable more detai­led ana­ly­sis by the authorities.

We loo­ked at medi­cines of major the­ra­peu­tic inter­est (MMTI) for which the ANSM drew up the first offi­cial list at the end of 2024, cove­ring around 10,000 pro­ducts (editor’s note : the same mole­cule can give rise to seve­ral pro­ducts if there are seve­ral dosages or manu­fac­tu­rers) out of the 17,000 marketed.

From mid-2016 to 2017, the dai­ly num­ber of MMTIs in short sup­ply remai­ned rela­ti­ve­ly stable, at around 100 drugs. It began to rise at the end of 2017, pea­king at around 250 during the first lock­down in 2020. From 2021 onwards, the upward trend acce­le­ra­ted, with a very clear peak emer­ging in the win­ter of 2022–2023, rea­ching 800 MMTI shor­tages per day. Since then, there has been a slow decline to 400 by the end of 2024. Howe­ver, this level remains very high com­pa­red to pre-Covid levels. The ave­rage num­ber of the­ra­peu­tic alter­na­tives avai­lable, without dis­rup­tion or risk of dis­rup­tion, was 5 in sum­mer 2021, 4 in Janua­ry 2022, 1.6 throu­ghout 2023, fal­ling to 2 at the end of 2024.

A stock shor­tage decla­red by a labo­ra­to­ry, across all drugs, leads to an ave­rage 11% drop in deli­ve­ries to phar­ma­cies throu­ghout the dura­tion of the dis­rup­tion. This is signi­fi­cant, of course, but we can­not talk about a wides­pread collapse.

Howe­ver, this ave­rage masks a wide range of situa­tions. As might be expec­ted, long stock shor­tages (more than four months) can lead to very signi­fi­cant declines in sales, of more than 75%, over seve­ral months. Conver­se­ly, shor­ter shor­tages of three months have no impact on sales. We can assume that these situa­tions reflect cases where who­le­sa­lers still had stocks or where manu­fac­tu­rers pro­du­ced the medi­cines in ques­tion on a just-in-time basis.

In March 2023, at the height of the 2022–2023 win­ter cri­sis, shor­tages led to 8 mil­lion fewer boxes being sold in phar­ma­cies. The ove­rall mar­ket for the medi­cines concer­ned was esti­ma­ted at 80 mil­lion or 120 mil­lion boxes per month, depen­ding on whe­ther MMTIs are defi­ned accor­ding to the ANSM or accor­ding to the defi­ni­tion that has been used by labo­ra­to­ries for seve­ral years, which adds seve­ral dozen para­ce­ta­mol-based medi­cines to the ANSM list. Depen­ding on the defi­ni­tion used, this repre­sents a 10% or 6.5% drop in the num­ber of boxes sold. Howe­ver, these figures do not consi­der any increases in the pro­duc­tion and sale of alter­na­tive treat­ments, which were not affec­ted by the shortage.

Bet­ween the end of 2022 and the end of 2024, labo­ra­to­ries built up an ave­rage of three weeks’ worth of stock for pro­ducts repor­ted as at risk of shor­tage. But des­pite this posi­tive trend, there are signs that these risks are beco­ming more com­plex. The num­ber of drugs sub­ject to ANSM mea­sures (import autho­ri­sa­tions, quo­tas on the num­ber of boxes or patient prio­ri­ti­sa­tion, for example) rose by ten points in 2024.

Dis­rup­tion reports alone are not very infor­ma­tive : when repor­ting a dis­rup­tion, labo­ra­to­ries must select a rea­son for the manu­fac­tu­ring issue they are expe­rien­cing from a pre­de­fi­ned list, but in most cases, the expla­na­tions pro­vi­ded are too gene­ral to iden­ti­fy a spe­ci­fic cause : “insuf­fi­cient pro­duc­tion capa­ci­ty”, “increase in sales volume”, or even “other…

Howe­ver, when taken toge­ther, two fac­tors point to sys­te­mic causes : the almost per­fect cor­re­la­tion bet­ween the change in the num­ber of medi­cines over time in rela­tion to the infla­tion curve, and the fact that all cate­go­ries of medi­cines have been affec­ted by the increase in the num­ber of shor­tages, even if they pea­ked at dif­ferent times.

One pos­si­bi­li­ty is the well-docu­men­ted increase in glo­bal demand for drugs, with a par­ti­cu­lar rise in pur­chases out­side the Uni­ted States and Europe. The conse­quences of the Covid cri­sis and the war in Ukraine have also cau­sed infla­tion to rise and dis­rup­ted pro­duc­tion chains world­wide, which may have affec­ted the drug mar­ket. Sup­ply chain issues for raw mate­rials, which are some­times sus­pec­ted of playing a signi­fi­cant role, do not appear to be among the main causes. In fact, labo­ra­to­ries only repor­ted this type of pro­blem in one in ten cases of shortages.

The stu­dy does not allow us to set­tle this debate. It would be neces­sa­ry to observe the actual price from one coun­try to ano­ther, which is not pos­sible at present because the dis­counts nego­tia­ted by indi­vi­dual govern­ments are not known. At this stage, the pre­cise causes of the shor­tages remain dif­fi­cult to pin­point, as they are neces­sa­ri­ly mul­ti­fac­to­rial and vary great­ly from one medi­cine to ano­ther. In gene­ral, it can be noted that the impro­ve­ment in the situa­tion since win­ter 2022–2023 has occur­red without any signi­fi­cant change in medi­cine prices in France. This sug­gests that stock shor­tages are not enti­re­ly deter­mi­ned by the issue of prices.

Interview by Anne Orliac

Stu­dy : “Ten­sions et rup­tures de stock de médi­ca­ments décla­rées par les indus­triels : quelle ampleur, quelles consé­quences sur les ventes aux offi­cines ?”