Traumatologisk forskning Anders Troelsen A-kursus, Traumatologi, Odense, September 2013
Hvorfor forskning? Hvilken behandlingsstrategi er bedst? Hvilket resultat kan forventes? Hvilke komplikationer er der risiko for? Hvilket implantat skal vi vælge? Hvilken teknik skal vi vælge?
Evidens Hvilket studiedesign giver den største styrke af evidensen ved sammenligning af to behandlinger?
Evidens Hvad sikres ved randomiseringen (fordeling ved lodtrækning)?
Graduering af evidens Publikationstype Evidens Styrke Metaanalyse, systemisk oversigt over RCT Randomiseret, kontrolleret studie Kontrolleret, ikke-randomiseret studie Kohorteundersøgelse Diagnostisk test (direkte metode) Case-kontrol undersøgelse Diagnostisk test (indirekte metode) Beslutningsanalyse Deskriptiv undersøgelse Mindre serier, oversigtsartikler Ekspertvurdering, ledende artikel Pedersen T et al. Hvad er evidensbaseret medicin? Ugeskrift for læger 2001; 163:3769-72. Ia Ib IIa IIb III IV A B C D
Operation eller konservativ behandling? Kvinde, 27 år, rask, alm. aktivitetsniveau
Operation eller konservativ behandling? Mand, 65 år, ryger, alm. aktivitetsniveau
Hvad skal vi bedømme klinisk og videnskabeligt? Helingsraten (non-union)? Funktionelt resultat? Andre komplikationer?
Hvilket / hvilke studie design egner sig bedst til at bedømme non-union raten?
Non-union ved konservativ behandling Non-union: < 1 % i historiske materialer fra 1960 erne Neer 1960 (2235 ptt.), Rowe 1968 (566 ptt.) Non-union: ca. 15 % i nyere metaanalyse Zlowodzki 2005 (2144 ptt.) Non-union ved operativ behandling Non-union: 0-3 % med nyere implantater Mckee 2007, Ferran 2010, Fridberg 2011
Hvilken behandling foretrækkes nu vurderet ud fra non-union rater? Cirka hvor mange skal opereres for at undgå 1 non-union? Findes der en måde at reducere NNT (numbers needed to treat)?
Khan 2009 fra: Robinson 2004, Brinker 2005
Andre signifikante risiko faktorer: Rygning + Co-morbiditet Murray, EFORT 2011 (1097 ptt.)
Hvordan skal vi vurdere det funktionelle resultat? Studie design? Validerede scoresystemer? (ca. 50 til skulder lidelser)
Litteratursøgning: Total number of publications Medline (Pubmed) search N=1369 Algorithm of search strategy results Potentially relevant publications that were evaluated according to title and abstract. N=30 Publications excluded: N=1339 Reason - Not in English N= 13 - Not relevant design N=852 - Not in English and not relevant design N= 474 Publications excluded: N=17 Reason - Not relevant topic N=16 - Children or adolescents included N=1 Publications for evaluation on design and endpoints N=13 Publications included N=5 COT society 2007 13, Ferran 2010 14,, Smekal 2008 16, Kulshrestha 2011 17 *Mirzatolooei 2011 18 Publications excluded: N=9 Reason - Not relevant topic N= 1 - Not relevant target population N= 1 - Children or adolescents included N=1 - Not relevant endpoints N=4
Studierne: Study Method No patients ( / ) LTF Mean age Fracture Interventions Endpoints* COT society. 2007 RCT 132 (87/24) 21 35,5 Completely displaced Internal plate fixation (N=67) or simple sling (N=65). Constant and DASH score. Union and complication rates Smekal et al. 2008 RCT 68 (52/8) 8 37,7 Completely displaced Intramedullary fixation (N=33) or simple sling (N=35). Constant and DASH score. Union and complication rates. Mirzatolooei 2011 RCT 60 (41/9) 10 33,4 Completely displaced and comminuted Internal plate fixation (N=31) or simple sling (N=29). Constant and DASH score. Union and complication rates Kulshrestha et al. 2011 pcs 73 (67/6) 5 32 Completely displaced Internal plate fixation (N=45) or simple sling (N=28). Constant score. Union and complication rates. RCT: randomized clinical trial, pcs: prospective cohort study. LTF: patients lost to follow-up * Only endpoints relevant to this article are mentioned. Fractures further subclassified.
Funktionelle resultater: Study Constant score Difference in score* Follow-up time (months) Other functional assessments COT society. 2007 Significant higher Constant score in plate group compared to simple sling group. 6 12 Results of DASH score correlates with results of Constant score. Mirzatolooei 2011 Significant higher Constant score in plate groupe compared to simple sling groupe. 11 12 Results of DASH score correlats with results of Constant score. Kulshrestha et al. 2011 Significant higher Constant score in plate groupe compared to simple sling groupe. 5 18 * difference in Constant score at the end of follow-up.
Signifikant bedre funktion!! Skal der opereres nu? Mckee 2007, COTS multicenter RCT studie
Hvad betyder en forskel på mellem 5 og 11 point? Hvad er den mindste kliniske relevante forskel i Constant score? Yian 2005 (1619 ptt., subjektivt raske)
Andre resultater skal inddrages: Symptomatisk malunion: Op til 30-40 % ved konservativ behandling Hill 1997, Nowak 2004 Skinnen fjernes pga. gener: Ca. 30 % ved moderne vinkelstabile skinner Fridberg 2011
Take Home Message Vi holder ryggen fri med evidensbaseret behandling Følgende skal især vurderes: Studie design (kend styrken) Hvilke end-points er relevante Vær kritisk i forhold til konklusioner