Are We “Squeezing The Balloon” When Reducing the Risk of Occupational Infection?

Reply to
Pan et al. •
Author(s): Ignacio Galed Placed , MD
Source: Infection Control and Hospital Epidemiology, Vol. 29, No. 1 (January 2008), pp. 95-96
Published by: The University of Chicago Press on behalf of The Society for Healthcare Epidemiology
of America
Stable URL: http://www.jstor.org/stable/10.1086/524914 .
Accessed: 15/05/2014 07:58

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 of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].

The University of Chicago Press and The Society for Healthcare Epidemiology of America are collaborating
with JSTOR to digitize, preserve and extend access to Infection Control and Hospital Epidemiology.

http://www.jstor.org

This content downloaded from 91.229.248.66 on Thu, 15 May 2014 07:58:02 AM

All use subject to JSTOR Terms and Conditions
 infection control and hospital epidemiology january 2008, vol. 29, no. 1

letters to the editor

Surgical Site Infections After the 2004 National Nosocomial Infection Surveillance (NNIS)
system data showed 3.23 infections per 100 cholecystectomies
Laparoscopic and Open Cholecystectomies in US hospitals for patients with the highest NNIS risk index
in Community Hospitals score.8 We commend Biscione et al.5 for providing valuable
data on cholecystectomies in community hospitals. However,
the results from the study do not appear to reflect the practice
in most US hospitals.
t o t h e e d i t o r — Laparoscopic cholecystectomy has be- We analyzed data collected by the Duke Infection Control
come the preferred method of performing gallbladder surgery Outreach Network on the rate of SSI following cholecystec-
over the past decade. Laparoscopic cholecystectomies are as- tomy to determine whether our findings were the same or
sociated with shorter hospital stay and convalescence, less dissimilar to those reported by Biscione et al.5 Data on the
pain and scarring, and lower rates of postoperative surgical risk of SSI were prospectively collected from 38,232 patients
site infection (SSI) than open cholecystectomies.1-4 The fact who underwent cholecystectomy in 31 community hospitals
that laparoscopic cholecystectomies are associated with fewer in the southeastern United States during the period from 1991
SSIs intuitively makes sense: laparoscopy access ports are to 2007. The median size of the participating hospitals was
short in length and only a fraction of the length of the incision 220 beds (range, 39-537 beds). We determined the rates of
used in open laparotomy. Biscione et al.5 reported in the SSI for laparoscopic and open cholecystectomies and then
September issue of the journal that patients who underwent stratified the rates by age, NNIS risk index score, American
laparoscopic cholecystectomy had lower rates of incisional Society of Anesthesiologists classification, wound class, and
SSI than patients who underwent open cholecystectomy; the year that surgery was performed. Hypothesis testing was
however, rates of deep incisional/organ space infections were performed with the x2 test for categorical variables, 2-tailed
similar in these 2 groups.5 The latter finding differs from our P values, and an a of .05.
experience and from the results reported in the existing lit- A total of 145 SSIs were identified following 38,232 cho-
erature, for several reasons. lecystectomies (ie, 0.38 infections per 100 procedures) during
Although 80% of all cholecystectomies in the United States the 15-year study period (Table); 35,316 (94%) of these cho-
were performed laparoscopically in 2002,6 only 59% of cho- lecystectomies were performed laparoscopically. The median
lecystectomies were performed laparoscopically in the Bis- patient age was 49 years. Open cholecystectomy was per-
cione et al. study.5 Furthermore, their results showed that the formed more often for older patients: the proportion of
use of laparoscopic cholecystectomy decreased over time, patients older than 60 years of age was 47% in the open
declining from 62% of procedures in 1996-1999 to 57% in cholecystectomy group and 29% in the laparoscopic chole-
2000 and later. The authors offered no explanation for this cystectomy group (P ! .001). Open cholecystectomy was per-
decline. We suspect that the criteria for choosing open or formed more often during the 1990s than it was after 2000:
laparoscopic cholecystectomy for the patients selected for this from 1991 to 2000, a total of 75% of cholecystectomies were
study were different from the criteria used in current practice performed laparoscopically; 94% of all procedures were done
in the United States. However, it is also possible that the laparoscopically after 2000 (P ! .001).
higher proportion of open procedures performed in the study The SSI rate was 1.37 infections per 100 open cholecys-
by Biscione et al.5 was the result of high rate of conversion tectomies and 0.30 infections per 100 laparoscopic procedures
from laparoscopic to open procedures as a result of technical (relative rate ratio, 4.61 [95% confidence interval, 4.29-4.97];
difficulties. P ! .001). We also performed subgroup analyses after strat-
The study by Biscione et al.5 was probably underpowered ifying the data by NNIS risk index score, age (older than 60
to assess whether the rate of deep SSI was the same or different years), American Society of Anesthesiologists classification,
in patients undergoing open procedures and patients under- wound class, surgical duration (more than 120 minutes or
going laparoscopic procedures. The authors examined 5,848 120 minutes or less), and the year in which the operation
procedures and observed a total of 39 deep incisional or organ was performed (before or after 2000). Laparoscopic chole-
space infections. In contrast, a study that showed a reduced cystectomies consistently had lower rates of SSI, compared
rate of deep incisional or organ space infection following with open procedures in all subgroups; with significant P
laparoscopic surgery examined 10 times that number of pro- values for each analysis, except for one comparison (Table).
cedures (54,504) and observed almost 7 times as many deep The type of surgery (laparoscopic or open) remained signif-
incisional/organ space infections (294).7 icantly associated with SSI in a multivariable model that con-
Finally, the high overall SSI rate in the study by Biscione trolled for all significant variables in the univariate analysis.
et al.5 restricted the generalizability of their findings. Biscione For the reasons cited above, we think that it is unwise to
et al.5 reported 3.57 infections per 100 procedures, whereas generalize the findings reported by Biscione et al.5 to patients

This content downloaded from 91.229.248.66 on Thu, 15 May 2014 07:58:02 AM

All use subject to JSTOR Terms and Conditions
 letters to the editor 93

table Rates and Relative Rate Ratios for Surgical Site Infection (SSI) Following Open and Laparoscopic
Cholecystectomies in a Cohort of 38,232 Patients from 31 Community Hospitals
Open procedure Laparoscopic procedure
(n p 2,916) (n p 35,316)
Relative rate ratio
Variable SSIs/proceduresa SSI rateb SSIs/proceduresa SSI rateb (95% CI) P
Overall 40/2,916 1.37 105/35,316 0.30 4.61 (3.21– 6.62) !.001
Patient characteristic
NNIS risk index score
0 6/1,067 0.56 44/22,926 0.19 2.93 (1.25-6.86) !.001
1 15/1,357 1.11 49/11,087 0.44 2.50 (1.41-4.45) !.001
x2 19/492 3.86 12/1,303 0.92 4.19 (2.05-8.57) !.001
Age
X60 years 12/1,551 0.77 54/25,052 0.22 3.59 (1.92-6.70) !.001
160 years 28/1,365 2.05 51/10,264 0.50 4.12 (2.61-6.52) !.001
ASA classification
!3 14/1,585 0.88 52/24,820 0.21 4.21 (2.34-7.59) !.001
x3 25/1,331 1.95 53/10,496 0.50 3.87 (2.42-6.16) !.001
Surgical duration
X120 minutes 20/2,069 0.97 91/33,107 0.27 3.52 (2.17-5.69) !.001
1120 minutes 20/847 2.36 14/2,209 0.63 3.73 (1.89-7.34) !.001
Wound class
CL-CC 32/2,762 1.16 100/34,444 0.29 3.99 (2.69-5.93) !.001
CO-I 8/154 5.19 5/872 0.57 9.06 (3-27.3) !.001
Time of surgery
Before 2000 6/894 0.67 10/2,648 0.38 1.78 (0.65-4.88) .26
After 2000 34/2,022 1.68 95/32,668 0.29 5.78 (3.92-8.53) !.001

note. ASA, American Society of Anesthesiologists physical status; CC, clean contaminated; CL, clean; CO, contam-
inated; I, infected; NNIS, National Nosocomial Infection Surveillance system.
a
No. of SSIs / no. of procedures.
b
No. of infections per 100 procedures.

in the United States. In fact, it may not be possible to use Address reprint requests to Luke F. Chen, MBBS, FRACP, 2100 Erwin
data on the risk of SSI following cholecystectomy that were Road, Durham, NC, 27710 ([email protected]).
Infect Control Hosp Epidemiol 2008; 29:92-94
collected prior to 2000 to assess the risk of SSI for patients 䉷 2007 by The Society for Healthcare Epidemiology of America. All rights
undergoing this procedure now. Finally, because in modern reserved. 0899-823X/2008/2901-0019$15.00.DOI: 10.1086/524335
practice the majority of open cholecystectomies have been
converted from attempted laproscopic procedures immedi-
ately after they are determined to have been unsuccessful, the
re f e re n c e s
risk of complications following open cholecystectomy is
likely to be quite different than was observed in the recent 1. Legorreta AP, Silber JH, Costantino GN, Kobylinski RW, Zatz SL. In-
past (eg, prior to 2000). However, our review of the risk of creased cholecystectomy rate after the introduction of laparoscopic cho-
lecystectomy. JAMA 1993; 270:1429-1432.
SSI in a cohort of 38,232 patients who underwent chole-
2. Hendolin HI, Paakonen ME, Alhava EM, Tarvainen R, Kemppinen T,
cystectomy supports the findings of other investigators who Lahtinen P. Laparoscopic or open cholecystectomy: a prospective ran-
have concluded that laparoscopic cholecystectomy is associ- domised trial to compare postoperative pain, pulmonary function, and
ated with significantly lower rate of SSI, compared with open stress response. Eur J Surg 2000; 166:394-399.
procedures.3,7,9,10 3. Zacks SL, Sandler RS, Rutledge R, Brown RS Jr. A population-based
cohort study comparing laparoscopic cholecystectomy and open cho-
lecystectomy. Am J Gastroenterol 2002; 97:334-340.
Luke F. Chen, MBBS (Hons), FRACP; 4. Inoue H, Itoh K-i, Hori H, et al. The cosmetic benefit of three-port
Deverick J. Anderson, MD, MPH; laparoscopic cholecystectomy and umbilical trocar insertion. Dig Endosc
Matthew G. Hartwig, MD; 1994; 6:49-51.
Keith S. Kaye, MD, MPH; 5. Biscione FM, Couto RC, Pedrosa TM, Neto MC. Comparison of the risk
of surgical site infection after laparoscopic cholecystectomy and open
Daniel J. Sexton, MD
cholecystectomy. Infect Control Hosp Epidemiol 2007; 28:1103-1106.
6. Livingston EH, Rege RV. A nationwide study of conversion from lapa-
roscopic to open cholecystectomy. Am J Surg 2004; 188:205-211.
7. Richards C, Edwards J, Culver D, Emori TG, Tolson J, Gaynes R. Does
From the Division of Infectious Diseases, Duke Infection Control Outreach using a laparoscopic approach to cholecystectomy decrease the risk of
Network (L.F.C., D.J.A., K.S.K., D.J.S.), and the Department of Surgery surgical site infection? Ann Surg 2003; 237:358-362.
(M.G.H.), Duke University Medical Center, Durham, North Carolina. 8. National Nosocomial Infections Surveillance (NNIS) System Report, data

This content downloaded from 91.229.248.66 on Thu, 15 May 2014 07:58:02 AM

All use subject to JSTOR Terms and Conditions
94 infection control and hospital epidemiology january 2008, vol. 29, no. 1

summary from January 1992 through June 2004, issued October 2004. assumption that laparoscopic cholecystectomy is cost effec-
Am J Infect Control 2004; 32:470-485.
tive, laparoscopic cholecystectomy savings may not overcome
9. Chuang SC, Lee KT, Chang WT, et al. Risk factors for wound infection
after cholecystectomy. J Formos Med Assoc 2004; 103:607-612. the increased consumption of healthcare resources resulting
10. Williams LF, Jr., Chapman WC, Bonau RA, McGee EC, Jr., Boyd RW, from the well-known rise in the number of cholecystectomies
Jacobs JK. Comparison of laparoscopic cholecystectomy with open cho- performed and in hospital use after laparoscopic cholecys-
lecystectomy in a single center. Am J Surg 1993; 165:459-465. tectomy is introduced.6-8 It is possible that all of these con-
cerns may have limited the availability of laparoscopy in our
institutions to some extent.
That our study might have been underpowered to detect
Reply to Chen et al. a lower risk of organ space infections after laparoscopic cho-
lecystectomy was acknowledged in the Discussion section of
our article.2 Chen et al.1 declined to assess the risk of organ
space infection in their study, so the assertion that our finding
t o t h e e d i t o r — We really appreciate the interest of Dr. differs from their experience is not supported by their data.
Chen and colleagues1 in our recent article.2 Several factors It is misleading to consider the study by Richards et al.9 as
may explain the lower rate of laparoscopic cholecystectomy valid evidence of reduced risk of organ space infection fol-
in our institutions. It is not the consequence of higher rates lowing laparoscopic cholecystectomy. The lower crude rate
of conversion to open cholecystectomy, because we kept con- of organ space infection after laparoscopic cholecystectomy
verted procedures in their original surgical group. First, note reported in that study was an unadjusted comparison and,
that our institutions adopted the National Nosocomial In- therefore, was not mentioned as a major finding by the au-
fection Surveillance (NNIS) system’s definition of operative thors.9 In that study, patients who underwent laparoscopic
procedure, which excludes outpatient (same-day) surgical cholecystectomy were obviously less prone to infection than
procedures.3 Thus, same-day laparoscopic cholecystectomies patients who underwent open cholecystectomy, because la-
were excluded from the analysis, lowering the overall pro- paroscopic cholecystectomies were shorter in duration, and
portion of laparoscopic procedures. Furthermore, our in- patients who underwent these procedures were younger and
creasing use of same-day laparoscopic cholecystectomy in re- less likely to have an American Society of Anesthesiologists
cent years, similar to increases in the use of this procedure classification of 3 or greater, dirty or contaminated wounds,
elsewhere around the world, can account for the apparent emergency procedures, or multiple procedures performed
reduction in the proportion of laparoscopic cholecystectomy through the same incision.9 Therefore, laparoscopic chole-
observed in our series over time. Second, the use of mini- cystectomy and open cholecystectomy could not be straight-
incision cholecystectomy as an alternative approach to both forwardly compared without a multivariate analysis, which
open cholecystectomy and laparoscopic cholecystectomy is was only provided for the overall risk of surgical site infection
rising in frequency in our country. For surveillance purposes, (SSI).9 We still believe that, if our extensive cohort of 5,848
we consider classic laparotomic cholecystectomies and mini- patients failed to show a difference in the risk of organ space
incision cholecystectomies to be open cholecystectomies. So, infection, a potential difference seems to be clinically irrel-
inclusion of data on procedures with mini-incision access evant in our setting.
would lower the rate of laparoscopic cholecystectomy but Chen et al.1 should note that more caution in needed when
leave the rate of open cholecystectomy unchanged. comparing the SSI rates we reported2 with those of the NNIS
Third, recall that our institutions are private and not uni- system. The most obvious reason for the higher SSI rate in
versity affiliated, and major differences exist in clinical prac- our cohort was the higher proportion of open cholecystec-
tices in comparison to public or university-affiliated hospitals. tomies. Second, the NNIS system emphasizes that, for SSI
Although economic concerns are obviously not unique to the rate comparisons to be valid, it must be known whether case
private sector, the economic pressures exerted by health plans finding included infections detected after discharge and the
assume a clear role in private institutions, particularly in a postdischarge surveillance method must be the same10; how-
resource-constrained country such as ours. Economic eval- ever, their own reports repeatedly fail to declare whether post-
uations conducted in developed countries have so far favored discharge surveillance was actually used and what proportion
laparoscopic cholecystectomy over open cholecystectomy, but of patients were reached by postdischarge surveillance.11 Chen
the cost savings of laparoscopic cholecystectomy may be less et al.1 also did not mention whether postdischarge surveil-
apparent in other types of institutions4 or when compared lance was used in their study. Therefore, any comparison
with minilaparotomy cholecystectomy.5 Moreover, some ways between their SSI rates and ours is irrelevant, and inferences
in which laparoscopic cholecystectomy has demonstrated about the generalizability of our data based on such com-
clear superiority over open cholecystectomy, such as shorter parisons are hasty and inaccurate. In our study, 55.5% of all
associated sick leave, do not necessarily represent a clear ben- SSIs were detected after discharge, even though a modest
efit to health plans and, accordingly, do not enter in their 49.5% of patients were reached by postdischarge surveillance.2
cost-effectiveness equations. More importantly, even with the In the study by Richards et al.,9 only 28.8% of SSIs were

This content downloaded from 91.229.248.66 on Thu, 15 May 2014 07:58:02 AM

All use subject to JSTOR Terms and Conditions
 letters to the editor 95

detected after discharge, suggesting that the proportion of 7. Legorreta AP, Silber JH, Costantino GN, Kobylinski RW, Zatz SL. In-
patients reached by out-of-hospital surveillance in that study, creased cholecystectomy rate after the introduction of laparoscopic cho-
lecystectomy. JAMA 1993; 270:1429-1432.
and in NNIS hospitals as a whole, is even less than that in 8. McMahon AJ, Fischbacher CM, Frame SH, MacLeod MC. Impact of
our study. What renders our practice incomparable to that laparoscopic cholecystectomy: a population-based study. Lancet 2000;
of most hospitals in the United States is, first of all, irrec- 356:1632-1637.
oncilable differences in surveillance methods. 9. Richards C, Edwards J, Culver D, Emori TG, Tolson J, Gaynes R. Does
For these reasons, we agree with Chen et al.1 that our results using a laparoscopic approach to cholecystectomy decrease the risk of
surgical site infection? National Nosocomial Infections Surveillance Sys-
must not be generalized to patients in the United States, and tem, Centers for Disease Control and Prevention. Ann Surg 2003; 237:
to the same extent, the results of US studies must not be 358-362.
generalized to our patients. It is not clear why they suggest 10. Mangram AJ, Horan TC, Pearson ML, Silver LC, Jarvis WR. Guideline
that a study could be generalized to a population other than for prevention of surgical site infection, 1999. Hospital Infection Control
that from which the sample was retrieved. Practices Advisory Committee. Infect Control Hosp Epidemiol 1999; 20:
250-278.
Chen et al.1 should explain further why it may not be 11. National Nosocomial Infections Surveillance (NNIS) system report, data
possible to use data on the risk of SSI collected prior to 2000 summary from January 1992 through June 2004, issued October 2004.
in a cohort study with concurrent controls. This concern may Am J Infect Control 2004; 32:470-485.
be more relevant for a study using historical controls. In- 12. Ishizaki Y, Miwa K, Yoshimoto J, Sugo H, Kawasaki S. Conversion of
creases in conversion rates are mainly driven by changes in elective laparoscopic to open cholecystectomy between 1993 and 2004.
Br J Surg 2006; 93:987-991.
the spectrum of indications for laparoscopic cholecystec-
tomy.12 The inclusion of the year of surgery as explanatory
variable was the only possible means to address this in our
data. Are We “Squeezing The Balloon” When
acknowledgments Reducing the Risk of Occupational
Financial support. F.M.B. reports receiving grant support N§0151041 from
Infection? Reply to Pan et al.
Capes/CNPq- IEL Nacional-Brasil for the original manuscript.
Potential conflicts of interest. The author reports no conflicts of interest
relevant to this article.
t o t h e e d i t o r — Under a witty title, Pan et al.1 described
Fernando M. Biscione, MD, MSc a needlestick injury that occurred while a cytopathologist was
performing fine-needle aspiration cytology (FNAC) using the
From the Health Sciences Postgraduate Course, Medicine High School, modified method that I and my colleagues proposed2 to elim-
Minas Gerais Federal University, Belo Horizonte, Minas Gerais, Brazil. inate the needle manipulation involved in classical FNAC.
Address reprint requests to Fernando M. Biscione, MD, MSc, 190 Alfredo They relate that the needle shot out when the cytopathologist
Balena av., Suite 7003, Santa Efigênia, Belo Horizonte, Minas Gerais, Brazil,
depressed the plunger to extract the material obtained by
31.130-100 ([email protected]).
Infect Control Hosp Epidemiol 2008; 29:94-95 aspiration, and they suggest a possible cause: the needle was
䉷 2007 by The Society for Healthcare Epidemiology of America. All rights clogged by a colloidal clot aspirated from a thyroid nodule.
reserved. 0899-823X/2008/2901-0020$15.00.DOI: 10.1086/524912 Unfortunately, the needle bounced and lodged in the cyto-
pathologist’s scalp. With no other argumentation, they as-
sume that clot aspiration is inherent to the modified method
ref e re n c e s of FNAC and so is a new injury risk associated with use of
the modified method.
1. Chen LF, Anderson DJ, Hartwig MG, Kaye, KS, Sexton DJ. Surgical site I agree that a clot might have clogged the needle and caused
infections after laparoscopic and open cholecystectomies in community
hospitals. Infect Control Hosp Epidemiol 2008; 29:92-94 (in this issue).
the accident. Nevertheless, I don’t know the physical law that
2. Biscione FM, Couto RC, Pedrosa TM, Neto MC. Comparison of the risk explains why aspiration of a clot is only possible when using
of surgical site infection after laparoscopic cholecystectomy and open the modified FNAC technique and not when using the con-
cholecystectomy. Infect Control Hosp Epidemiol 2007; 28:1103-1106. ventional FNAC method. Also, the reason for which the nee-
3. Horan TC, Emori TG. Definitions of key terms used in the NNIS system.
dle’s odd flight path and its unlucky ending must be consid-
Am J Infect Control 1997; 25:112-116.
4. Teerawattananon Y, Mugford M. Is it worth offering a routine laparos- ered an exclusive result of the modified FNAC method eludes
copic cholecystectomy in developing countries? A Thailand case study. me. My personal experience includes more than 6,500 FNAC
Cost Eff Resour Alloc 2005; 3:10. procedures, more than 3,500 of which were performed with
5. Nilsson E, Ros A, Rahmqvist M, Backman K, Carlsson P. Cholecystec- the modified method. Although I cannot determine the exact
tomy: costs and health-related quality of life: a comparison of two tech-
number, I have had a few experiences with both classical and
niques. Int J Qual Health Care 2004; 16:473-482.
6. Hobbs MS, Mai Q, Fletcher DR, Ridout SC, Knuiman MW. Impact of modified FNAC in which the needle shot out suddenly when
laparoscopic cholecystectomy on hospital utilization. ANZ J Surg 2004; I depressed the plunger. Fortunately, no injury ever followed.
74:222-228. Finally, I don’t see the line of argument by which Pan et

This content downloaded from 91.229.248.66 on Thu, 15 May 2014 07:58:02 AM

All use subject to JSTOR Terms and Conditions
96 infection control and hospital epidemiology january 2008, vol. 29, no. 1

al.1 conclude that reporting a diagnostic method that, with Anatomia Patologica, Seccion de Citologia, Hospital Juan Canalejo, Xubias
no loss of effectiveness, permits us to eliminate needle ma- de Arriba 84, La Coruna, Spain, ([email protected]).
Infect Control Hosp Epidemiol 2008; 29:95-96
nipulation and hence, to reduce the risk of injury, is equiv- 䉷 2007 by The Society for Healthcare Epidemiology of America. All rights
alent to “squeezing the balloon.” Again, I am in perfect agree- reserved. 0899-823X/2008/2901-0021$15.00. DOI: 10.1086/524914
ment with Pan et al.1 when they recommend that we use our
heads to perform FNAC safely. However, and to avoid mis-
understandings, I would dare to specify their advice further,
adding that, if possible, we shouldn’t use our scalps, but we re f e re n c e s
should make use of our common sense. 1. Pan A, Signorini L, Magri S, De Carli G. Scalp needlestick injury during
fine-needle aspiration cytologic evaluation without needle manipulation:
Ignacio Galed Placed, MD William Tell in the laboratory, not quite. Infect Control Hosp Epidemiol
2006; 27:996.
2. Galed-Placed I, Pertega-Diaz S, Pita-Fernandez S, Vazquez-Martul E. Fine
From the Servicio de Anatomia Patologica, Seccion de Citologia, Hospital needle aspiration cytology without needle manipulation to reduce the risk
Juan Canalejo, La Coruna, Spain. of occupational infection in healthcare personnel. Infect Control Hosp
Address reprint requests to Ignacio Galed Placed, MD, Servicio de Epidemiol 2005; 26:336.

This content downloaded from 91.229.248.66 on Thu, 15 May 2014 07:58:02 AM

All use subject to JSTOR Terms and Conditions