Operations Research in Hospitals-Diagnosis and Prognosis. by David H. Stimson; Ruth H.StimsonReview by: John LuckmanJournal of the Royal Statistical Society. Series C (Applied Statistics), Vol. 23, No. 3 (1974), pp.442-443Published by: Wiley for the Royal Statistical SocietyStable URL: http://www.jstor.org/stable/2347142 .

Accessed: 17/12/2014 12:31

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

.

Wiley and Royal Statistical Society are collaborating with JSTOR to digitize, preserve and extend access toJournal of the Royal Statistical Society. Series C (Applied Statistics).

http://www.jstor.org

This content downloaded from 128.235.251.160 on Wed, 17 Dec 2014 12:31:59 PMAll use subject to JSTOR Terms and Conditions

442 APPLIED STATISTICS book after such a long interval. It is a fascinating account of the work of the Operational Research Section at Coastal Command, a section which included two Nobel prize winners and five Fellows of the Royal Society, which demonstrates how the idea that "scientific methods of analysis might give useful assistance to effective executive action" was turned into a decisive activity in the war against the U-boat.

It offers a view of OR in its early phases of development in which its general character is perhaps more clear than it appears today. Mathematical and statistical analyses abound, but there are none of the elaborate techniques that have been developed over the last two decades or so. Insistence on systematic and scientific observations and recording of data together with logical analyses and critical questioning of criteria provided the foundation for advice-often in the face of considerable adverse opinion. The results achieved were outstanding by any standards-perhaps the most impressive being the increase in U-boats destroyed per attack from 2 per cent in 1941 to over 40 per cent in 1944.

The book has chapters on the various aspects of the war against the U-boat: the organi- zation of flying effort, including maintenance, inspection, scheduling, comparison of air- craft types, length of operational tours; navigation and weather problems; the use of radar; visual problems, sighting, flying heights, aircraft camouflage; principles of depth charge attacks, aiming errors, lethality under different conditions; air defence of convoys and offensive operations against the U-boat.

For anyone teaching or practising OR or wishing to know more about OR and its origins, or just interested in the development of tactics and strategy in the war against the U-boat, I wholeheartedly recommend this book.

R. HARRIS Institute for Operational Research

Operations Research in Hospitals-Diagnosis and Prognosis. By David H. Stimson and Ruth H. Stimson. Chicago, Illinois Hospital Research and Educational Trust, 1972. $4 95.

Although dealing mostly with the American scene, the majority of the diagnoses put forward by the authors as to why there are difficulties in implementing operations research studies will be just as familiar to OR scientists and statisticians on this side of the Atlantic.

The authors begin by reviewing hospital OR studies to be found in the literature, wisely ignoring those using an industrial engineering approach which is much favoured in the U.S. The literature is divided into seven categories covering such topics as staffing, admission/discharge and inventory control. The first category is the old chestnut, schedul- ing in outpatients clinics. The authors argue that this problem area, like many of the others covered in this section does not lack a body of theory since queuing theory results have been available for years. Neither is there any lack of researchers capable of putting forward perfectly valid alternative solutions since the veriest OR novice in the hospital field must have looked at the Outpatient Department at some early stage. What is lacking is implementation and, in Section 3, the authors piece together the difficulties. Looking at it from the hospital administrator's point of view, they list and discuss thirteen such difficulties.

The three I would select as being most important are: -the tendency to view the hospital as a separable system; -the tendency to view the hospital as a mechanistic system; -the omission of variables that are difficult to quantify. The discussion of these shortcomings is brief but very much to the point, dealing with

such items as the lack of more holistic models of what is a highly interdependent system of activities, the lack of recognition of a hospital as a complex and sensitive social system and the difficulty of measuring the outputs of the service. All these points are backed up by

This content downloaded from 128.235.251.160 on Wed, 17 Dec 2014 12:31:59 PMAll use subject to JSTOR Terms and Conditions

BOOK REVIEWS 443

a bibliography of considerable length, albeit somewhat patchy in content. The authors would have done the reader a great service if the bibliography had been more selective.

Section 4 explores the researcher/hospital administrator relationship and posits some ways in which the relationship could be changed so as to make it more effective. Having argued forcibly for the direct involvement of both researcher and administrator in any study, the authors suggest firstly that "Change in Education" would be beneficial. American experience is that both sides are being more widely exposed to the others viewpoints and this can only be for the good.

The second suggestion, "Changes in the Ways OR Services are Provided to Hospitals" is necessarily linked to the U.S. problem of a fragmented health service with small, indepen- dent organizations. We should count ourselves fortunate that regional OR teams can so easily be formed in this country although it is not certain that we make the best use of those we have.

The third and final suggestion "Development of a Meta-Discipline" comes as rather an anti-climax. Defining a meta-disciplinary approach as one which emphasizes the bringing of persons together because of their common interest in a particular subject (while inter- disciplinary implies bringing a diversity of disciplines to examine a particular subject), the authors are in effect repeating the concept of involvement of the participators in the health care system in solving their own problems, assisted where necessary by specialists in various fields. I would have hoped that the best specialists, be they OR men, statisticians, economists or social scientists, would already have adopted a stance of research by collaboration, but the catalogue of woeful experiences in the previous chapters is a frightening reminder that those ideals are not always met.

All researchers in the hospital and health service field should buy this book, keep it open at the page listing the 13 difficulties and resolve to overcome as many of them as possible in their next research project.

JOHN LUCKMAN Institute for Oper-ational Research

Bibliography on Experiment and Treatment Design Pre-1968. By Walter T. Federer and Leslie N. Balaam. Edinburgh, Oliver and Boyd, 1972. 769 p. 10". ?9.

In preparing this bibliography, in the specific area of experiment and treatment design. over 100 journals, most of which are listed in the introduction, were completely scanned and many others were partially scanned. The result is a publication containing in excess of 8,500 different references offering virtually complete coverage of the literature prior to 1968. The individually numbered citations are presented as five sub-bibliographies in Sections VI-X of the book, alphabetical-chronological ordering being used within sections. Section VI is an expanded and revised version of Federer's pre-1950 literature review (Math. Res. Center Tech. Summary Report No. 405, Wisconsin, 1964) and Section VII is, with some additions, the Bibliography of Experiment Design 1950-1967 previously released by the authors in 1970 as Report No. 1080 at Wisconsin. A bibliography for treatment design covering the 1950-67 period is given in Section VIII which, to make it self-contained, repeats some of the entries given in the previous section. Section IX is on topics related to the conduct and analysis of experiment and treatment design. Amongst the many topics included in this section are outliers, transformations, computer programs for analyses of experiments, covariance and multivariate analyses, whilst, for example, X2 analyses and variance component analyses are excluded. Twenty-seven textbooks on experiment and treatment design are referenced in Section X, all but two (Japanese) accompanied by their original listing of contents.

To aid the user of the bibliography the authors have classified the references in the following manner. Each paper, according to its main theme, is first classified as: experi- ment design; treatment design; analysis; sequential and fixed sample size procedures;

16

This content downloaded from 128.235.251.160 on Wed, 17 Dec 2014 12:31:59 PMAll use subject to JSTOR Terms and Conditions