Accuracy and cost-effectiveness of two clinical decision rules for the...


Update content

Title Accuracy and cost-effectiveness of two clinical decision rules for the diagnosis of deep vein thrombosis and pulmonary embolism in residents of nursing homes and homes for the elderly.
Period 08 / 2008 - 01 / 2013
Status Current
Research number OND1334467
Data Supplier ZonMw Projectenpoort


This study aims to quantify in residents of nursing homes and homes for the elderly suspected of DVT or PE, the accuracy and cost-effectiveness of two existing diagnostic rules for DVT and PE that were yet developed and validated in primary and secondary care respectively (see below and systematic review). The primary outcome will be the safety (i.e. no VTE within 3 months follow up) in the low risk groups (low probability of DVT and PE) defined by the corresponding scores. Hence the analysis focuses on the estimation of the negative predictive value of both rules, the negative likelihood ratio, and the sensitivity. Besides the rules ability to exclude VTE, we also focus on their ability to correctly estimate the risk of DVT and PE and thus their potential to detect (timely) the presence of these disorders (aiming to reduce potential under-diagnosis). We will also study whether both rules are accurate as such or whether they require updating or improvement to enhance their safety and accuracy in the setting of nursing homes and homes for the elderly. The cost-effectiveness of both strategies will be addressed in the entire group of nursing home patients suspected of VTE. Also, reasons for decisions by the physicians against the rules, i.e. not referring a patient if the rule indicates a high risk and referring even though the rule indicates a low risk, will be studied.
Background. Residents of nursing homes and homes for the elderly are significantly more at risk of venous thromboembolism (VTE), i.e. deep vein thrombosis (DVT) or pulmonary embolism (PE). Moreover, in the elderly diagnosis of VTE is more difficult due to more obscured presentation and presence of comorbidities. This may lead to unnecessary referrals for objective diagnostic testing (overdiagnosis). But also to underdetection of VTE presence and thus under-referral (as it is also considered too burdening in many cases). Diagnostic decision rules for diagnosing DVT and PE in patients suspected of VTE have been developed and validated. The rules use signs and symptoms in combination with D-dimer testing. The rules have never been validated in the high risk (elderly) population of nursing homes and homes for the elderly.Objective. To quantify the accuracy and cost-effectiveness of two decision rules for VTE in residents of nursing homes and homes for the elderly.Study design. Prospective diagnostic study in nursing homes and homes for elderly with a 3-month follow-up.Study population. Residents of nursing homes and homes for the elderly with a suspicion of DVT or PE.Measurements. All patients undergo systematic history taking, physical examination, and point-of-care D-dimer testing. The existing diagnostic decision rules for DVT or for PE will be applied. Patients with low risk of having DVT or PE will (in principle) not be referred for further diagnostic work-up and withheld from anticoagulant therapy. All other patients will be referred for diagnostic work-up, i.e. compression ultrasound for DVT and (spiral CT) angiography or VQ-scanning for PE.Primary endpoints. 3-months incidence of VTE, defined as DVT and non-fatal or fatal PE. Power/data analysis. For DVT, expecting a failure rate of 1% in the low score (risk) group and to exclude a rate of 3% or more we need about 335 patients in the low score category (alpha=0.05 one-sided as any percentage <1% is desired, beta = 0.2), requiring about 1200 patients suspected of DVT. For PE, expecting a failure rate of 0.5% in the low risk group and to exclude 2%, we need about 340 patients in the low risk group, and approximately 950 patients suspected of PE. Economic evaluation. Using a Markov decision model, costs and health effects when applying the rules will be weighted against the costs and effects of the current situation. Time schedule. 2 months preparation, 30 months inclusion, 4 months to complete analysis and reporting.

Abstract (NL)

Achtergrond. Bewoners van verpleeg- en verzorgingshuizen hebben een hoger risico op veneuze tromboembolie (VTE):diep veneuze trombose (DVT) of longembolie (LE). Diagnostiek van VTE bij ouderen is vaak lastig door aanwezige comorbiditeiten en niet eenduidig klachten en verschijnselen. Dit leidt enerzijds tot onnodig verwijzen van patienten voor aanvullende diagnostiek (overdiagnostiek) en anderzijds tot onderdiagnostiek (-behandeling) in geval van niet of te laat gedetecteerde VTE. Voor secundaire en primaire care patienten met verdenking op DVT of LE, zijn klinische beslisregels ontwikkeld voor de diagnostiek van beide aandoeningen. Beide regels gebruiken gegevens van anamnese, lichamelijk onderzoek en de D-dimeer test. De regels zijn echter nog nooit gevalideerd bij hoogrisico patiënten uit verpleeg- en verzorgingshuizen.Doel. Kwantificeren van de accuratesse en kosteneffectiviteit van twee beslisregels (voor DVT en LE) bij bewoners van verpleeg- en verzorgingshuizen.Onderzoeksopzet. Prospectief diagnostisch onderzoek met drie maanden follow-up.Onderzoekspopulatie. Bewoners van verpleeg- en verzorgingshuizen met verdenking op DVT of LE.Metingen. Bij alle patiënten wordt gestructureerd een anamnese en lichamelijk onderzoek uitgevoerd plus een point-of-care D-dimeer test. De beslisregel voor DVT of LE (afhankelijk van verdenking) wordt toegepast. Patiënten in de laag risico groep volgens de regel, worden (in principe) niet verwezen voor aanvullende diagnostiek en niet behandeld met anticoagulantia. Alle andere patiënten worden verwezen voor aanvullende diagnostiek:compressie echo-onderzoek van extremiteiten voor DVT, en (spiraal CT) angiografie en/of VQ-perfusie scanning voor LE. Primaire eindpunten. 3 maanden incidentie van VTE. Power/analyse. DVT: uitgaande van verwachte 1% DVT in laag-risicogroep is en om 3% te kunnen uitluiten met alfa=0,05 (eenzijdig) en bèta=0.2, zijn 335 patiënten in laag-risicogroep nodig en 1200 met verdenking DVT. LE: uitgaande van verwachte 0.5% in laag-risicogroep en om 2% uit te kunnen sluiten zijn 340 patiënten nodig in deze groep en 950 met verdenking LE.Economische evaluatie. Met een Markov beslissingsmodel worden de kosten en effecten als gevolg van het gebruik van de regel gewogen tegen die van de huidige zorg. Tijdpad: 2 mnd voorbereiding, 30 mnd inclusie en follow-up, 4 mnd data analyse en rapportage.

Related organisations

Related people

Project leader Prof.dr. K.G.M. Moons

Go to page top
Go back to contents
Go back to site navigation