Research Policy 46 (2017) 591–604 Contents lists available at ScienceDirect Research Policy jo ur nal ho me page: www.elsevier.com/locate/respol The impact of health research on length of stay in Spanish public hospitals Antonio García-Romero a,∗ , Álvaro Escribano b , Josep A. Tribó c a IE University – IE Business School, c/María de Molina, 31-Bis, 28006 Madrid, Spain b Department of Economics, Universidad Carlos III de Madrid, c/Madrid, 126, 28903 Getafe Madrid, Spain c Department of Business, Universidad Carlos III de Madrid, c/Madrid, 126 28903 Getafe Madrid, Spain a r t i c l e i n f o Article history: Received 1 November 2015 Received in revised form 31 December 2016 Accepted 9 January 2017 Available online 27 January 2017 Keywords: Medical Surgical and basic research Clinical outcomes Length of stay Hospital efficiency Absorptive capacity Translational approach a b s t r a c t Research is a key determinant of health improvement. However, there is little empirical evidence show- ing how the research conducted in hospitals affects healthcare outcomes. To address this issue, we used panel data of 189 Spanish public hospitals over the period 1996–2009 to estimate the causal effect of both clinical and basic research on hospitals’ efficiency, measuring their impact on the average length of stay (LOS). We considered two fixed effects econometric models; one for medical and the other for surgical specialties respectively. Our results show that increases in the quantity of research produced in medical (surgical) disciplines contribute significantly to the reduction of hospital LOS in medical (surgi- cal) specialties. This effect is greater for hospitals with higher absorptive capacity (high R&D investment and with teaching status). There is also clear evidence that basic research produces efficiency gains in clinical outcomes. Furthermore, we have identified other important determinants of hospitals’ efficiency namely, hospitals’ characteristics, human resources, diagnostic activity, hospital investment and hos- pitals’ absorptive capacity. Finally, we evaluated the economic impact of increases in medical, surgical and basic research on hospitals’ cost efficiency gains by measuring the corresponding reduction in the average cost of stay in Spanish hospitals. © 2017 Elsevier B.V. All rights reserved. 1. Introduction To date, much attention has been paid to the evaluation of scientific research on health from the point of view of measur- ing the performance of scientific output (i.e., number of published papers) or scientific impact (i.e., number of citations, citations per document). This approach has played a pivotal role in the design of research policies or in establishing priority settings in health research. Nevertheless, in recent years, research policy has leaned towards a more stakeholder-oriented view that emphasizes the societal returns of research (Bornmann, 2013). Consistently, there is a need for new approaches capable of measuring the “real” effects of research on society (Smith, 2001; Cozzens, 2000; Patel et al., 2011). In the case of health-related research, its societal impacts can be observed in many aspects such as the development and improve- ment of new drugs and health technologies (Toole, 2007), or the creation of spin-offs (Haeussler and Colyvas, 2011; Quilter-Pinner ∗ Corresponding author. E-mail addresses: antonio.garcia@ie.edu, agr22@faculty.ie.edu (A. García-Romero), alvaroe@eco.uc3m.es (Á. Escribano), joatribo@emp.uc3m.es (J.A. Tribó). and Muir, 2015). However, one of the most relevant returns from research conducted in hospitals is the improvement of clinical out- comes such as a reduction in length of stay (henceforth LOS), which has both direct and indirect effects on cost reduction and other efficiency gains. In this context, the main aim of this paper is to explain the effects of research carried out in hospitals on their average LOS, and subsequently, on hospital costs. Formally, LOS is the term used to measure the duration of a single episode of hospitalization (Preedy and Watson, 2010). There are many indicators commonly used for measuring clinical outcomes. However, there is a lack of consen- sus among experts as to what constitutes a healthcare outcome (Porter, 2010). Our decision to choose LOS as the dependent vari- able to be explained by our models is based on three reasons: (i) LOS is a standard measure commonly used in health care management and is available for any hospital in any country; (ii) this character- istic favors comparability among different studies; (iii) LOS is an outcome that can be translated in terms of economic cost, which allows extracting policy recommendations. We describe the mechanisms through which research out- put improves healthcare outcomes taking advantage of two approaches: the translational research approach (direct channel) and the absorptive capacity approach (indirect channel). On the http://dx.doi.org/10.1016/j.respol.2017.01.006 0048-7333/© 2017 Elsevier B.V. All rights reserved.