The title of the article provides some idea about the study carried out; it could be improved if contained progesterone as this has been the main factor being studied. The authors who participated in this study had very much related abilities and qualified very well to carry out this study. A large list of authors accommodates researchers, epidemiologists, biostatisticians, obstetricians, gynecologists, and experts in cancer research. The presence of this learned group of people guaranteed a very powerful study. This study was funded by National Cancer Institute. There is a list of some authors who had affiliations with some pharmaceutical companies but no conflict of interest was started by any of these authors or could be identified.
Problem Statement: estrogen alone for hormone replacement therapy (HRT) in postmenopausal women places them at increased risk of endometrial hyperplasia maturing to endometrial cancer. Past research work has also shown that combination therapy (estrogen and progesterone) can have better results with little effect on the endometrial response towards hyperplasic changes. But some studies have provided contradictory results even with combination therapy.
Purpose of the study: to enhance understanding of HRT and to verify the potential role of progesterone in reducing the effects of unopposed estrogens on the endometrium.
The knowledge base regarding the current understanding of the topic is very much related to the nursing practice for managing their related patients.
Next to rationale, the objectives of any research activity play an integral role. In fact, objectives are the sole reason for conducting any research work. At times objectives are developed by the researcher but their presentation may not be appropriate. A good objective should be a SMART (specific, measurable, achievable, relevant, and time-bound) one. The authors of this article have not been able to describe the objectives properly and clearly.
One of the important components of any experimental design is the randomization of study subjects to different groups (Agresti & Finlay, 1999; Rothman & Greenland, 1998).In research, randomized controlled trials (RTC) are at the top for generating any evidence followed by prospective and then retrospective studies (case-control). It seems very odd when authors have explicitly mentioned two RTCs on the issue explored and after that carried out a case-control study to improve their understanding of evidence base on the topic.
An extensive literature review has been carried out by the authors to substantiate their rationale for conducting this study. The existing evidence provides the information available on the topic. There is a recognizable body of research evidence is present on the topic. The majority of the studies referred to in this article support the role of combination HRT in minimizing the unopposed effect of estrogen on endometrial hyperplasia. A few of the studies concluded in contradiction to these findings which prepared a base for conducting this research study. Use of current evidence or from the recent past makes the rationale a stronger argument. It also suggests that no research activity has been carried out similar to the proposed one in between this period and this proposed one is very well justified. But in this article, the majority of the reference cited have been older than five years. Some are even more than ten years old.
Ethical considerations are the hallmark of any research; its importance increases even further when there is any external factor is introduced to humans or animals (Rothman & Greenland, 1998). The authors did mention about ethical review bodies of various hospitals but they did not explicitly mention any approval obtained for conducting this study.
The use of hormones was taken as an outcome variable in this article. Multiple logistic regression models were built. There were different categories of the outcome variable (unopposed use of estrogen, combined HRT, continuous HRT) used in different models. The hypothesis of the study was: use of combined HRT (including progesterone with estrogen) will have a protective effect on the endometrial mucosa.
In the methodology section, the authors have described the sampling of the study participants in a very detailed and comprehensive way. Potential eligible cases were: Black or White women in the age range of 50-79 years, who were residents of contiguous nine-county Philadelphia, Pennsylvania at the time of diagnosis. They were newly diagnosed with endometrial cancer between July 1, 1999, and June 30, 2002. These cases were identified through an active surveillance system at 61 hospitals. Medical record and cancer registry, Pennsylvania Cancer Registry provided the information on these cases.
Controls were selected through random digit dialing from the same areas as those of the cases. These controls were matched on race (Black or White) and age (5 years age groups). These controls did not have endometrial cancer or hysterectomy performed. Both the cases and c controls had a household telephone number, were living in a non-institutional setting, we’re able to speak English, and did not have any cognitive problem. The controls were selected through a survey research firm. Women with endometrial cancer were excluded from the cases if the interval the diagnosis and case ascertainment was more than 18 months, or the interval between ascertainment and contact for the screening interview was more than 12 months.
The authors did not mention any estimation of sample size or any target number of cases or controls for this study. The team could identify 1,185 incident cases through various procedures but the final analysis was carried out on 511 postmenopausal endometrial cancer cases. The controls were matched on race and age and final analysis was completed on 1,412 controls. The study participants were divided into two groups, cases, and controls through matching. The cases were identified through a cancer registry and six hospitals in Philadelphia while controls were matched from the same geographic communities from where cases were located. The setting fulfilled the requirements for utilization in this study.
No prior pilot study was done. This study was carried out utilizing the study design of a case-control study. Although previous studies’ findings and the existing evidence were used in building the rationale for this study.
Before starting the interview, eligible participants were sent a letter that shared with them the purpose of the study and their role in the study. It was followed by an initial telephone call to screen the participants after getting their consent and confirming the eligibility of the participants.
The authors have not mentioned anything related to the development of data collection tools. Also, the article does not contain any information on pretesting of the questionnaires or methodology to be used as well as no statement regarding the reliability or validity of the study tools has been described. Although the trained interviewers were hired to carry out this study they did not know anything about the hypothesis of the study and had no information on the risk (increased or decreased) of any medication/hormone. Interviews were carried out by telephone and took almost one hour from completion. Study participants used photos for responding to the interviewer’s questions. No procedure has been described which suggested any care taken to secure the confidentiality of the participants and talked about the rights of the participants.
Study variables have been described in greater detail to understand definitions of different terms used in the article. These definitions are important in order to remove the ambiguity which may produce as well as keeping all the concerned persons; authors, analysts, reviewers, readers, and other audience, at the same level of understanding.
The authors have given a detailed account of statistical analysis. The y have mentioned the software used in this analysis along with its version. Different categories formed of various variables and the type of analysis performed also has been stated. Data management is a group of activities conducted right from the time the data collection instruments are developed and it continues through the stages of data collection, validation, editing to entry, and analysis. For any valid research study, the role of the data management process is vital. If there is any breakage in this chain or disturbance of quality then the whole of the study is out at stake. The authors of this study did not mention anything related to data management except information on analysis.
The result section has been dealt with with great care. Authors have shared almost all important results obtained form this study. There were 511 cases and 1, 1412 controls. Cases were a bit older than controls. White, never married, and Jews were more in the cases group as compared to the controls. Some of the co-morbidities like: diabetes, cholelitiasis and cholecystits were more common in cases. The frequency of HRT use by postmenopausal women came out as 44.5 percent (42.3 percent in cases and 45.3 percent in controls) and this use was much higher in White women as compared to Blacks. As compared to never users of HRT, women with unopposed use of estrogen for 3 or more years reflected twice the risk of doubling endometrial cancer.
The authors performed conditional logistic regression analysis as it was a matched case-control study. They constructed various models to explore different dimensions of this study. They used p-value as well as 95 percent confidence intervals (95 % CI) to compare different groups. They also described looking at the confounders.
Discussion section of any scientific article makes the article acceptable or not! It is this section where results of the current study are compared to the previous ones and thus a detail debate is carried out. The authors try to bring their study in the line of existing evidence if the results are alike; at times the results are against what had been presented in earlier studies then a plausible explanation is required by the authors to state his or her view point. If the results take an absolutely new turn then the authors also suggest further research in the area.The authors have provided detailed discussion on the study carried out. They concluded that the current study proved reduced risk when a combined estrogen/progesterone replacement was used. They have compared their study with the previous ones and found matching results. However, they did not give any suggestion for further or future research in this are or a similar area to explore other unanswered issues.
The authors have discussed the limitations this study had. These have been described as selection and detection bias. The major reasons have been refusals from the attending physicians or the eligible candidates themselves for joining the study. The authors have rejected the possibility of any misclassification bas or recall bias by arguing through various aspects.
These findings are important as far as nursing practice is concerned. They provide an insight in the association of different types of hormonal preparation for treating osteoporosis especially in postmenopausal women. A combination of these hormones success in reducing the risk of endometrial cancer will improve understanding of nurses on this issue.
References
Agresti A., & Finlay B. (1999). Statistical Methods for the Social Sciences. 3rd edition, pp 21. Prentice Hall, New Jersey.
Rothman KJ., & Greenland S. (1998). Modern Epidemiology. 2nd edition, pp. 67-70. Lippincott – Raven, Philadelphia.
The Practice of leadership, Web.