Home | About CytoJournalEditorial Board | Archived articles | Search CytoJ Articles | Subscribe | Peer review policies | CytoJournal Quiz Cases
  Reviewer corner | Author corner | OA Steward’s corner | CF member’s corner | Join as CF member | Manuscript submission | Open Access (OA) Advocacy
CytoJournal All 'FULL TEXT' in HTML are FREE under "open access" charter of CytoJournal.
To login for downloading any PDF OR to request TOC (Table of Content) by e-mail, please click here
Home Email this page Print this page Small font size Default font size Increase font size Cytopathology Foundation
Navigate here
 » Next article
 » Previous article 
 » Browse articles
Resource links
 »  Similar in PUBMED
 »  Search Pubmed for
 »  Search in Google Scholar for
 »  Article in PDF (507 KB)
 »  Citation Manager
 »  Access Statistics
 »  Reader Comments
 »  Email Alert *
 »  Add to My List *
* Registration required (free)  

  In this article
 »  Competing Intere...
 »  Authorship State...
 »  References
 »  Article Tables

 Article Access Statistics
    PDF Downloaded386    
    Comments [Add]    
    Cited by others 5    

Recommend this journal


CytoJournal 2009,  6:12

Test group biases and ethical concerns mar New England Journal of Medicine articles promoting HPV screening for cervical cancer in rural India

Department of Pathology, Magee-Womens Hospital of University of Pittsburgh Medical Center, 300 Halket Street, Pittsburgh, PA 15213, USA

Date of Submission16-Apr-2009
Date of Acceptance03-Jun-2009
Date of Web Publication16-Jul-2009

Correspondence Address:
R Marshall Austin
Department of Pathology, Magee-Womens Hospital of University of Pittsburgh Medical Center, 300 Halket Street, Pittsburgh, PA 15213
Login to access the Email id

Source of Support: None, Conflict of Interest: None

PMID: 19633722

Rights and PermissionsRights and Permissions

How to cite this article:
Austin R M, Zhao C. Test group biases and ethical concerns mar New England Journal of Medicine articles promoting HPV screening for cervical cancer in rural India. CytoJournal 2009;6:12

How to cite this URL:
Austin R M, Zhao C. Test group biases and ethical concerns mar New England Journal of Medicine articles promoting HPV screening for cervical cancer in rural India. CytoJournal [serial online] 2009 [cited 2019 Sep 22];6:12. Available from: http://www.cytojournal.com/text.asp?2009/6/1/12/53466

The study conclusions, released from prepublication embargo by the New England Journal of Medicine (NEJM) at 5:00 PM on April 1, 2009, were dramatic. "In a low resource setting, a single round of HPV testing was associated with a significant reduction in the numbers of advanced cancers or deaths from cervical cancer." Furthermore, the study abstract asserted, "No significant reductions in the numbers of associated deaths were observed in the cytology testing group or in the VIA group, as compared with the control group." [1]

Simultaneously, at 5:00 PM on April 1, 2009, a press release from Qiagen, a Netherlands Holding Company and a manufacturer of the digene HPV test, stated: "Landmark study in New England Journal of Medicine shows HPV testing significantly reduces deaths from cervical cancer, compared to other methods including Pap." [2] It is now known that the manufacturer received pre-embargo access to contents of the article and the accompanying editorial. The NEJM customer service office confirmed on April 14 that pre-embargo release was not from the journal.

A close look at the NEJM paper suggests that unexpected biases might have occurred in some of the test group arms of the study which sought to compare a single round of cervical cancer screening with HPV testing, cytology, and visual inspection with acetic acid (VIA) among over 131,746 "healthy women between the ages of 30 and 59 years" in rural India. In fact, the study acknowledged that the positive predictive value (PPV) for detecting CIN 2 - 3 was 19.3% in the cytologic testing group, higher than 11.3% in the HPV testing group, and study results indicate that essentially the same numbers of cervical cancers were detected after positive screening test results in the cytology arm (88) as in the HPV arm (87). Cervical cancer detection rates were 0.344% in the cytologic testing group and 0.321% in the HPV testing group [Table 1]. Among cervical cancers detected after a positive screening test, more cervical cancers were detected in the earliest and most favorable Stage (IA) in the cytology arm (58/88, 65.9%) than in the HPV arm (45/87, 51.7%), and there were fewer advanced (Stage II+) screening-detected cervical cancers in the cytology arm (10/88, 11.4%) than in the HPV arm (14/87, 16.1%) [Table 1]. The preliminary 2005 publication of study data had earlier reported that cytology had a higher detection rate for CIN 2 - 3 (1.0%) than for HPV testing (0.7%) and observed the following: "Our present findings indicate that the detection rates of HPV testing did not show any improvement over cytology. Furthermore, the currently available test (HC2) is expensive and requires a relatively sophisticated laboratory operation." [3] Only the individuals performing HPV testing are specifically described in the 2005 report as having been trained "intensively." What new information then led to the quite different study conclusions reported in the 2009 publication and the simultaneous manufacturer's press release?

The 2009 publication employed the well-recognized "intention-to-treat-principle" in which all eligible patients assigned to a study arm are included in the final analyses for each study arm, whether or not the patients adhere to the research protocol management strategy. [4] Therefore, the final study analyses included in each arm cervical cancer patients who had not been screened, 32 in the HPV arm and 42 in the cytology arm [Table 2]. Since there were 10 more unscreened cancer patients and 5 more cervical cancer deaths in the cytology arm than in the HPV arm and since these unscreened cancer patients predictably had higher rates of cervical cancer, more advanced stages, and less favorable outcomes, the net effect of the inclusion of the larger number of unscreened cervical cancer patients in the cytology arm was to make the cytology arm appear more like the unscreened control arm of the study. Of 54 cervical cancer-related deaths in the cytology cohort, 27 deaths were in the "assigned but not screened" group and another 18 deaths were in patients who had abnormal cytology results. For cervical cancer deaths in the HPV arm, 19% fewer deaths (22) were in the "assigned but not screened" group, and 33% fewer deaths (12) occurred in women with abnormal screening HPV test results. Given the higher PPV and specificity of cytology compared to HPV testing, the outcomes suggest that follow-up colposcopic examinations, tissue sampling procedures, and treatments were not equally effective in the two groups. Significantly inconsistent adherence to management strategies in different study arms may also introduce study bias in investigations, even in studies applying the intention-to-treat-principle. [4]

Other aspects of the study data also support the hypothesis that bias was introduced in the study group arms during treatment of patients who developed cervical cancer. For example, the death rate for clinically detected cervical cancer patients in the HPV test-negative group of the HPV arm was 0% as compared to 40.9% among clinically detected cancer patients in the cytology-negative group of the cytology arm [Table 3]. Remarkably, the 0% death rate occurred in the HPV group despite the fact that 62.5% of the HPV test-negative cervical cancers were "advanced stage" (Stage II+). Among patients with positive screening results, similar unexpected outcome differences are also evident [Table 1]. Even though the numbers of cervical cancers in both screen-detected groups were essentially the same (HPV 87; cytology 88), more favorable Stage IA screen-detected cancers were detected in the cytology arm (58) than in the HPV arm (45). Also, fewer unfavorable Stage II+ cancers were screen detected in the cytology arm (10) than in the HPV arm (14). Even after acknowledging the higher negative predictive value of HPV testing reflected in [Table 3], deaths in the HPV arm were unexpectedly one-third lower (12) than in the more stage-favorable cytology arm (18). The 2005 publication [3] states that "Women with suspected invasive cancer were referred to the surgical oncology department of NDMCH (Nargis Dutt Memorial Cancer Hospital) for investigations and treatment with surgery and radiation." The study also notes that "The most common form of treatment for invasive carcinoma in the intervention arm was radical hysterectomy followed by radical radiotherapy." No data are reported on the extent of similarities or differences in cancer treatments between the test group arms. The effect of the bias of unexpectedly favorable treatment outcomes in the HPV arm contributes to the study conclusion that HPV screening (rather than some other factor) preferentially prevented cervical cancer deaths. The NEJM report does acknowledge that "both practitioners and subjects were aware of the study group assignments." [1]

The NEJM publication reports that the study was supported by the Bill and Melinda Gates Foundation, through the Alliance for Cervical Cancer Prevention (ACCP). The paper also states that "no potential conflict of interest was reported." However, questions have been raised in the peer-reviewed literature as to whether or not the Gates Foundation and the ACCP might have in effect "become enamored with the promise of science and new technologies at the expense of delivering available preventives today". [5] Questions have also been raised about the partnership between AACP's coordinating organization and the HPV test manufacturer. [5-7] Similarly, the Gates Foundation was recently acknowledged as having funded a large project supporting publication of a special supplemental issue of the journal Vaccine entitled "Overview of Human Papillomavirus-Based and Other Novel Options for Cervical Cancer Screening in Developed and Developing Countries." [8] The HPV test manufacturer is also listed as providing corporate support, [8] acknowledged as "critical for meetings and travel." The two lead authors of the NEJM article and one of the NEJM editorial writers are specifically listed among the large number of international contributors. The presence of not even a single cytopathologist among the very large number of contributors supported for international travel and meetings clearly suggests an anticytology bias for the proceedings, one perhaps mirrored in the NEJM. The two selected authors of the accompanying NEJM editorial [9] also turn out to have co-authored no fewer than 34 PUBMED-cited publications with the former Vice President for Research and Development and Chief Scientific Officer of the HPV test manufacturer. Industry co-authorship was recently cited as the second most common conflict-of-interest in high impact journals such as NEJM. [10]

Although the NEJM study upon first inspection may be interpreted as applying primarily to low-resource settings where even very limited screening and diagnostic testing are difficult to support, the accompanying editorial, [9] press release [2] and media coverage all strive to suggest otherwise. According to the editorial writers, "The implications of the study in rural India are immediate and global." [9] According to the New York Times (NYT), the study indicates broadly that a "new DNA test outperforms the  Pap smear More Details." [11] Strangely absent from any of the commentary, however, has been any concern that the study included a "control" group of 31,506 "healthy women" who were allowed to largely go without any screening at all during the study, one approved by both local and international "institutional scientific and ethical committees." [3] Sixty-four women in the "control" group died of cervical cancer along with another 70 unscreened women "assigned to undergo screening," numbers of cervical cancer deaths substantially in excess of the number which occurred in widely criticized experiments conducted at New Zealand's Auckland National Women's Hospital in the latter half of the last century. [12] The deaths of so many unscreened study "control" subjects made the NYT-quoted comment by a Stanford gynecologist, "The study is another nail in the coffin for Pap smears," seem particularly insensitive and ironic. [11]

The unexpected increase in the presence of cervical cancer in both the unscreened cytology and VIA arms of the study was a key foundation for the authors' surprising statement that "no significant reductions in the numbers of advanced cancers or deaths were observed in the cytologic-testing group or in the VIA group, as compared with the control group." [1] When unscreened study cancer victims are included in the "control" group, it becomes clear that significant reductions in advanced stage cervical cancers occurred in virtually all screening group arms [Table 4]. The comparative reductions in advanced stage cervical cancer after a single round of (less expensive) conventional cytology are not significantly different from those achieved with a single round of (more expensive) HPV testing (P = 0.2655), even before taking into consideration other study biases. Furthermore, the study's cytology screeners were trained for only 3 months, one-fourth of the usual year-long training for cytotechnologists. This was almost certainly suboptimal to confront the challenging conditions in the rural Osmanabad district. Pap screening, when successfully implemented, reduces cervical cancer rates by 60 - 90% within 3 years of introduction to populations that have not been screened; these reductions in incidence and mortality are consistent and dramatic across populations. [5,13] Even a single round of screening with conventional cytology and limited training in rural India appears to have been remarkably effective, despite the published interpretations of study data in the NEJM. As noted by many of the same authors in their 2005 publication, "Our results clearly show that good quality cytology can be implemented even in a rural setting of a developing country with reasonable investment, while HPV testing does not give any better detection of CIN2/3 lesions, despite the higher investments." [3] Careful reading of the 2009 publication data and assessments of the numerous internal biases suggest that the widely reported conclusions may be significantly misleading.

 » Competing Interest Statement Made by all Authors Top

No competing interest to declare by any of the authors.

 » Authorship Statement by all Authors Top

All authors of this article declare that we qualify for authorship as defined by ICMJE http://www.icmje.org/#author.

Each author has participated sufficiently in the work and take public responsibility for appropriate portions of the content of this article.

Each author acknowledges that this final version was read and approved.

 » References Top

1.Sankaranarayanan R, Nene BM, Shastri SS, Jayant K, Muwonge R, Budukh AM, et al . HPV screening for cervical cancer in rural India. N Engl J Med 2009;360:1385-94.   Back to cited text no. 1    
2.Press Release. Source: Qiagen. Landmark study in the New England Journal of Medicine shows HPV testing significantly reduces deaths from cervical cancer, compared to other methods including Pap. Wednesday April 1, 2009, 5:00 PM EDT.   Back to cited text no. 2    
3.Sankaranarayanan R, Nene BM, Dinshaw KA, Mahe C, Jayant K, Shastri SS; et al . A cluster randomized controlled trial of visual, cytology, and human papillomavirus screening for cancer of the cervix in rural India. Int J Cancer 2005;116:617-23.   Back to cited text no. 3    
4.Montori VM, Guyatt GH. Intention-to-treat principle. CMAJ 2001;165:1339-41.  Back to cited text no. 4  [PUBMED]  [FULLTEXT]
5.Suba EJ, Donnelly AD, Furia LM, Huynh ML, Raab SS. Cervical cancer prevention for all the world′s women: genuine promise resides in skilled quality management rather than novel screening approaches. Diagn Cytopathol 2007;35:187-91.   Back to cited text no. 5    
6.Suba EJ, Murphy SK, Donnely AD, Furia LM, Huynh LM, Raab SS. Systems analysis of real worlds obstacles to successful cervical cancer prevention in developing countries. Am J Public Health 2006;96:480-7.  Back to cited text no. 6    
7.Suba EJ, Frable WJ, Raab SS. Cost-effectiveness of cervical cancer screening in five developing countries. N Enlg J Med 2006;354:1535-6.   Back to cited text no. 7    
8.Cuzick J, Arbyn M, Sankaranarayanan R, Tsu V, Ronco G, Mayrand MH, et al. Overview of human papillomavirus-based and other novel options for cervical cancer screening in developed and developing countries. Vaccine 2008;26S:K29-41.  Back to cited text no. 8    
9.Schiffman M, Wacholder S. From India to the world: a better way to prevent cervical cancer. N Engl J Med 2009;360:1453-5.  Back to cited text no. 9  [PUBMED]  [FULLTEXT]
10.Jaqsi R, Sheets N, Jankovic A, Motomura AR, Amarnath S, Ubel PA. Frequency, nature, effects and correlates of conflicts of interest in published Clinical Cancer Research. Cancer 2009;115:2783-91.  Back to cited text no. 10    
11.McNeil, DG. New DNA test outperforms Pap smear. New York Times. April 7, 2009.   Back to cited text no. 11    
12.McCredie MR, Sharples KJ, Paul C, Baranyai J, Medley G, Jones RW, et al. Natural history of cervical intraepithelial neoplasia and risk of invasive cancer in women with cervical intraepithelial neoplasia 3: a retrospective cohort study. Lancet Oncol 2008;9:425-34.   Back to cited text no. 12  [PUBMED]  [FULLTEXT]
13.Agency for Healthcare Research and Quality. Screening for cervical cancer: File inventory. Available from: http://www.ncbi.nlm.nih.gov/books/bv.fcgi?rid¼hstat3.section.4193. [accessed on 2009 Apr 14].  Back to cited text no. 13    


  [Table 1], [Table 2], [Table 3], [Table 4]

This article has been cited by
1 result 1 Document Advancing cervical cancer prevention in India: Implementation science priorities
Krishnan, S., Madsen, E., Porterfield, D., Varghese, B.
Oncologist. 2013;
2 Cervical cancer screening in Germany. Current status [Gynäkologische Krebs-vorsorge in Deutschland Gegenwärtiger Stand]
Schneider, V.
Pathologe. 2012; 33(4): 286-292
3 Lessons learned from successful Papanicolaou cytology cervical cancer prevention in the socialist republic of Vietnam
Suba, E.J., Raab, S.S.
Diagnostic Cytopathology. 2012; 40(4): 355-366
4 Cervical screening: Toward a new paradigm? [Un nouveau paradigme pour le dépistage du cancer du col utérin?]
Lavoué, V. and Bergeron, C. and Riethmuller, D. and Daraï, É. and Mergui, J.-L. and Baldauf, J.-J. and Gondry, J. and Douvier, S. and Lopès, P. and De Reilhac, P. and Quéreux, C. and Letombe, B. and Marchetta, J. and Boulanger, J.-C. and Levêque, J.
Journal de Gynecologie Obstetrique et Biologie de la Reproduction. 2010; 39(2): 102-115
5 Reply to: Austin et al. CytoJournal 2009;6:12 (Unfounded claims mar scientific critique)
Jeronimo, J. and Barone, M. and Luciani, S. and Lu, R. and Sherris, J. and Torod, J. and Tsu, V.
CytoJournal. 2009; 6(23)


Previous article Next article


  Site Map | Copyright and Disclaimer
© 2007 - CytoJournal | A journal by Cytopathology Foundation Inc with Wolters Kluwer - Medknow
New version online since 1st July '08
Open Access