martin haslberger

pharmacist, meta-researcher and medical student

[PLOS Medicine @PLOSMedicine]

In this meta-research study, Martin Holst and colleagues investigate changes to prespecified primary outcomes in the historic registries of clinical trials. @MartinHolst9 @m_haslberger @Strech_Da @LGHemkens @questbih @berlinnovation @ChariteBerlin https://t.co/pCpWzxcF6U

[Lars G. Hemkens @LGHemkens]

We analyzed 1746 trials completed 2009-17 at German university medical centers:
55% changed primary outcome
23% w major change

BUT:
14% of changes only visible in registry history-not when looking at last entry only
1% reported change

#PLOSMedicine
https://t.co/qdCfh2Jk8M
1/2

[Dr Marguerite O'Riordan @Marguerite_OR]

Hot off the press!
https://t.co/UygzV5l1vQ

[TranspariMED @TranspariMED]

One in five of all #oncology clinical trials (phase 3) changed primary outcomes

-> and most did not mention this in publications

https://t.co/StkJnatB87 @m_haslberger @Strech_Da @DeborahZarin @reshmagar @GhassanKaram @ISRCTN @ChrisW_PharmaG

[TranspariMED @TranspariMED]

Clinical trial transparency accelerates access to better treatments

- says NIHR's @LucyChappell2, who leads the funder with the strongest research waste safeguards worldwide:

https://t.co/Z698NdUYIz @NIHRresearch @tarik_suljic @m_haslberger

[TranspariMED @TranspariMED]

Adverse events during #clinicaltrials not reported in journals

-> trial registries provide crucial evidence

cc @m_haslberger @NDevito1 @LGHemkens @eturnermd1 @Mad_In_America @HRA_Latest @matt_westmore @meerpohl @cochranecollab

[martin haslberger @m_haslberger]

When a clinical trial is registered retrospectively, do authors report and justify this transparently in journal publications, as recommended by the ICMJE? We checked and the answer (unsurprisingly) is: rarely. With @Strech_Da, now published in @BMJ_Open: https://t.co/l0UYAUhnJo

[TranspariMED @TranspariMED]

Very mixed picture on research waste safeguards among #clinicaltrials funders in Europe.

Strong transparency policies in Sweden, Norway, France (@Inserm) and Belgium.

But in Italy and Spain, public funders do not even require trials to be registered... 4/n
@LGHemkens

[Martin Smollich @ErnMedBlog]

Aktuelles Beispiel für unseriöse Angstmacherei @mdrwissen: Die genannte Studie zeigt KEINEN Zusammenhang zw. #Aspartam & #Alzheimer. Sie zeigt: Sehr hohe Aspartam-Konz. (über Ernährung nie erreichbar!) führen an isolierten Neuroblastom-Zellen (!) zu oxidativem Stress. ...

[TranspariMED @TranspariMED]

41% of German university trials switched primary outcomes

14% had ‘hidden’ discrepancies

This can lead to biased or misleading conclusions for health care decision-making, argue @MartinHolst9 & @m_haslberger

https://t.co/DtVcR0KUbo #QRP @arzneitelegramm @aerztezeitung

[Lars G. Hemkens @LGHemkens]

"Hidden changes" of primary outcomes in trials are difficult: The primary outcome definition in the paper and in the latest/newest registry entry is identical, but when you dig deeper & look at older registry entries you see the outcome changed.

We found this in 1 of 7 trials…

[martin haslberger @m_haslberger]

Registrations are checked inadequately during peer review anyway, but editors/reviewers would have to dig through historical registration records to not miss a relevant number of primary outcome discrepancies.

[martin haslberger @m_haslberger]

New preprint with @MartinHolst9, @LGHemkens, @Strech_Da, et al., looking at primary outcome changes that are „hidden within the registry“, i.e., not apparent from comparing the latest reg. entry with the publication. https://t.co/sdjsiABHtR

[martin haslberger @m_haslberger]

We found that 23% of trials in our sample have primary outcome changes after study start reported within the registration history. 41% had changes between the latest entry and publication. 14% had changes that were “hidden behind the latest entry”.

[TranspariMED @TranspariMED]

"In contrast to ICMJE guidance, retrospective registration is disclosed and explained only in a small number of retrospectively registered studies"

https://t.co/YXgEqKGhlb @NaudetFlorian @ronaklmno @NDevito1 @jsross119 @eloder @GustavNilsonne