Comparison of Healthcare Resource Utilization (HcRU) between Age Groups of Women Diagnosed with Gynecological Cancers in the U.S

Godwin Okoye^1*, Henry Antwi¹ and Rachel vincent^2
*Correspondence: Godwin Okoye, Department of Pharmaceutical Health Services Research, University of maryland school of pharmacy, baltimore, MD, USA, Tel: 4439359693, Email:

Keywords

Healthcare Resource ● Utilization ● Uterine neoplasm ● Cervical neoplasm ● Ovarian neoplasm ● Menopausal

INTRODUCTION

According to Centers for Disease Control and Prevention (CDC), gynecological cancer (GC) refers to when malignant cells originate from the female reproductive organs. There are five major types: cervical, ovarian, uterine, vaginal, and vulvar cancers [1]. All women are at risk of developing one or more types of GC and the risk increases with age. Between 2012- 2016, about 94,000 women were diagnosed with gynecological cancer each year. It is interesting to note that Uterine cancer is the fourth most common cancer for women in the U.S with about 66% diagnosed at an early stage due to the abnormal vaginal bleeding that occurs as an early symptom [2]. Incidence rate for cervical cancer has dropped by more than 50% in the past 50 years due to screening exercises that can detect changes in the cervix before they become malignant [3]. There has been an upward trend in the incidence of ovarian cancer owing to the increasing use of oral contraceptives and hormonal therapy for menopausal women [4]. GC generally accounts for about 11.2% of all female deaths [5].

Over the years, it has been discovered that GC is more prevalent in older, mostly post-menopausal women [6]. This has gathered much attention in the scientific community, encouraging research that could improve health outcomes for older women with GC [7,8]. However, little research attention has focused on the clinical burden of GC among younger, mostly premenopausal women [9]. This research focuses on self-reported use of health care resource of younger women with GC when compared to older women to improve their health outcomes and general survivorship.

The main objective of this study was to compare healthcare use among two distinct age groups of women with GC. i.e., women most likely to be premenopausal and women most likely to be post-menopausal.

Materials and Methods

Study design and data source

A retrospective cross-sectional repeated measures study was conducted to compare all patient-reported demographics and HcRU related to GC in women most likely to be pre-menopausal and women most likely to be postmenopausal using the sample adult file of NHIS (National Health Interview Survey) data from 2015 to 2018. NHIS is the principal source of information on the health of the civilian noninstitutionalized population of the United States. It is one of the major data collection programs of the National Center for Health Statistics (NCHS) which is part of the Centers for Disease Control and Prevention (CDC). The main objective of NHIS is to monitor the health of the United States population through the collection and analysis of data on a broad range of health topics [10].

Study sample

We identified women who reported having a diagnosis of either of the following: Cervical, Ovarian or Uterine cancer. A few women in the cohort had more than one GC. In this study, women most likely to be premenopausal were defined as those who are less than 50 years old while women who are most likely to be post-menopausal were defined as those more than 50 years old.

Baseline demographics, clinical and behavioral characteristics

According to the systematic review conducted by Phillips, et al., we utilized the behavioural model of health services utilization as a framework [11]. we selected our covariates based on individual and population level factors of patients in the cohort. These include Risk factors (race, marital status, obesity, pregnancy [12-14] Enabling factors (health insurance coverage, region in the U.S [15,16], Behavioral factors (smoking, drinking [17].

Health care resource utilization

We focused on nine areas of HcRU which are: obstetrics and gynecological visit, specialist visit, general practitioner visit, hospitalization or in-patient admissions, emergency room visit, office-based services, home healthcare services, surgery and surgical procedures, preventive healthcare. Preventive health care focused on women in the cohort who have ever undergone pap smear or who have received HPV vaccination. The number of individuals who reported utilizing any of the above health service was recorded at baseline and proportions were also taken by age group.

Statistical analysis

Baseline characteristics and outcome measures of interest were obtained for individuals in the cohort. We reported count and column percentage for categorical variables. The covariates were: race, census region, marital status, obesity, smoking, alcohol, health insurance, and pregnancy. Comparison between defined age groups of the women in the cohort was done for baseline characteristics using the Pearson’s chi square (χ²) test.

For HcRU, values for the nine areas mentioned above was computed and stratified amongst age group. Pearson’s chi square (χ²) test was used as a measure of association. Annual trends in health care resource utilization Figure 1 was extrapolated to give an insight on the independent changes happening across each year within the study period.

A univariate logistic regression model was used to compare HcRU between women most likely to be pre-menopausal (defined as women < 50 years) and women most likely to be post-menopausal (defined as women >50 years) across nine areas of health services mentioned above. Only predictors with p<0.05 were utilized for multivariate logistic regression model.

All statistical analysis were performed using SAS v9.4 ((SAS Institute Inc, Cary, NC), and statistical significance was set at α = 0.05.

Classification of evidence

This study provides class IV evidence that women most likely to be premenopausal diagnosed with GC can use comparable health care resources with postmenopausal women diagnosed with GC despite GC being more prevalent in women most likely to be postmenopausal.

Results

Based on our data, Gynecological cancers were discovered in 1518 responses representative of 2633569 responses who reported having at least one kind of GC. We had 1047 women (1742470 weighted) who are most likely to be post-menopausal and 471 women (891099 weighted) who are most likely to be pre-menopausal. The following cases were found for the individual types of GC: 577 women reported having uterine cancer, 728 women with cervical cancers, and 323 ovarian cancers. The mean age and standard deviation of the women with GC was: Uterine (64 ± 13), Cervical (53 ± 15) and Ovarian (60 ± 15). Overall, the mean age of women in the cohort was (58 ± 16). Statistical significance at p<0.05 was achieved for behavioral factors in the study which includes alcohol consumption and smoking.

Based on this study, Table 1A and 1B displays the baseline characteristics of women in the cohort. Alcohol use and smoking was associated with the two age groups being compared at p<0.05. In simpler terms, pre-menopausal women were more likely to consume alcohol when compared to postmenopausal women with approximate values of 88% and 79.93% respectively. For smoking, the approximate values are 59.79% and 50.26%.

Health care resource utilization

Table 2 displays percentage values for HcRU of interest across age groups considered to be most likely premenopausal and most likely postmenopausal. Based on our data, Premenopausal women utilized more health care for Obstetrics and Gynecological visits (51.78% vs. 33.51%), Emergency room visits (41.29% vs. 31.19%), Hospitalization (30.59% vs. 21.9%) and Preventive care (46.33% vs. 25.72%). Surgery/surgical visits was the only HcRU parameter of interest that wasn’t statistically significant at p<0.05.

Figures 1(A-I) illustrated annual trends in HcRU. A clear distinction was marked for premenopausal women in the four areas mentioned above where they tend to use more healthcare than postmenopausal women. One notable outcome of this research was the upward trend in hospitalization for premenopausal women.

Table 3 displays results from the logistic regression models comparing premenopausal women with GC and postmenopausal women with GC. Out of nine parameters of interest, premenopausal women showed significant HcRu in four of the parameters. With the following results: Obstetrics and Gynecological visits (OR 2.38, (95% CI 1.89-2.97)), Emergency room visits (OR 1.73, (95% CI 1.39-2.17)), In-patient hospitalization (OR 1.92 (95% CI 1.51-2.43)), Preventive care (OR 2.695 (95% CI 2.14-3.39)). Surgery/surgical visits was not consistent with an association between premenopausal and postmenopausal women. Similar results were observed in the sensitivity analysis displayed in Table 4.

Discussion

This retrospective cross sectional repeated measures study used NHIS data from 2015-2018 to compare HcRU among women most likely to be premenopausal and women most likely to be post-menopausal who have had at least one diagnosis of ovarian cancer, uterine cancer, or cervical cancer.

From our baseline statistics, race wasn’t statistically significant when compared with both age groups, but it is known from previous studies that being white increases the risk of gynecologic cancer [18]. It is also interesting to note that both age groups had similarity in being obese which interesting because there is an association between obesity and age [19]. Behavioral factors like drinking and smoking were found to be strongly associated with younger women with gynecological cancer which isn’t surprising because of past literatures pointing out drinking and smoking as a coping mechanisms for young adult women [20].

To improve the understanding of longitudinal use patterns, we quantified HcRU changes using a combination of descriptive statistics and regression models. Overall, of nine essential area of health care use in gynecological cancer, premenopausal women were seen to significantly use preventive health care such as pap smear and HPV vaccination which is known to prevent cervical cancer or detect it at an earlier stage. This utilization pattern is consistent with a paper written by Sirovich, et al. [21]. Premenopausal women were also observed to have been more hospitalized. The trend for hospitalization was also observed to have had a continuous increase throughout the study period. Although younger women are expected to have lesser hospitalization due to expected less comorbidities of certain diseases that could be predictors of increased hospitalization [22]. Premenopausal women also had more frequent emergency room visits and visits to Obstetrics and Gynecologist maybe due to the fact that most of them are within reproductive age and are most likely to have a partner or be willing to have children as described by Schramm, et al. [23]. Surgery/surgical visits was not associated with the age group. This could be due to hysterectomy performed as a choice treatment for gynecologic cancer. Although in past literatures [24], younger women were seen to visit general practitioners more than older women, the reverse was the case for our current study.

Our study has several strengths, first we utilized the NHIS (National Health Interview Survey) which is the principal source of information on the health of civilian noninstitutionalized population in the United States. This survey data is able to capture the many demographic and socioeconomic characteristics of participants. It also finds its use in monitoring trends in illness and disability. To the best of our knowledge, this study is the first to make comparison between healthcare resource utilization of women who are most likely to be premenopausal and women most likely to be postmenopausal diagnosed with gynecological cancer in the U.S. As such, this could serve as the basis for other study to be conducted.

There are a few limitations to this study. First, some important variables were missing in the previous years. As a result of missing variables, our study was limited to four years. One notable inconsistency in covariates was use of chemotherapy and absence of specific procedures like hysterectomy. Pap smear and HPV vaccination was recorded in the 2018 data and with all the response missing. Second, given the nature of this study as self-reported questionnaire, it is subject to non-response and reporting bias. However, according to Rosenman, et al. [25], there is some agreement between selfreported questionnaires and medical records. Third, due to the short survival rate and late detection of some gynecological cancers, this population-based survey may have underestimated the prevalence.

Conclusion

This study presents some evidence that younger women who are most likely to be premenopausal utilize healthcare in a comparable way as older women who are most likely to be postmenopausal does. As such, research relating to gynecological cancer should not be limited to the most prevalent group.

References

1. Centers for Disease Control and Prevention. “Basic information about gynecologic cancers. Centers for Disease Control and Prevention.” (2021).
2. Tarver, Talicia. “Cancer facts & figures 2012. American cancer society (ACS)” (2012): 366-367.

Google Scholar, Crossref

3. “Cervical Cancer: Statistics.” Cancer.Net. (2022).
4. American Society of Clinical Oncology. “Ovarian, fallopian tube and peritoneal cancer: Statistics.” Cancer. Net. (2020).

Google Scholar

5. American Cancer Society. “Cancer facts & figures 2022.”
6. Sundar, Sudha, Richard D. Neal and Sean Kehoe. “Diagnosis of ovarian cancer.” BMJ 351 (2015).

Google Scholar, Crossref, Indexed at

7. Dumas, Lucy, Alistair Ring, John Butler and Tania Kalsi, et al. “Improving outcomes for older women with gynecological malignancies.” Cancer Treat Rev 50 (2016): 99-108.

Google Scholar, Crossref, Indexed at

8. “Retooling for an Aging America: Building the Health Care Workforce.”

Crossref, Indexed at

9. Mattsson, Elisabet, Lisa Ljungman, Kim Einhorn and Inger Sundström Poromaa, et al. “Perceptions of care after end-of-treatment among younger women with different gynecologic cancer diagnoses – a qualitative analysis of written responses submitted via a survey.” BMC Womens Health 20 (2020): 1-9.

Google Scholar, Crossref, Indexed at

10. Bureau, U. S. C. (2021, October 8). National Health Interview Survey (NHIS).
11. Phillips, Kathryn A., Kathleen R. Morrison, Ronald Andersen and Lu Ann Aday. “Understanding the context of healthcare utilization: assessing environmental and provider-related variables in the behavioral model of utilization.” Health Serv Res 33 (1998): 571.

Google Scholar, Indexed at

12. Kato, Ikuko, Suketami Tominaga, and Chikahiro Terao. “An epidemiological study on marital status and cancer incidence.” JPN J Cancer Res 80 (1989): 306-311.

Google Scholar, Crossref, Indexed at

13. Madison, Terri, David Schottenfeld, Sherman A. James and Ann G. Schwartz, et al. “Endometrial cancer: Socioeconomic status and racial/ethnic differences in stage at diagnosis, treatment, and survival.” Am J Public Health 94 (2004): 2104-2111.

Google Scholar, Crossref, Indexed at

14. Kalliala, Ilkka, Georgios Markozannes, Marc J. Gunter and Evangelos Paraskevaidis, et al. “Obesity and gynaecological and obstetric conditions: Umbrella review of the literature.” BMJ 359 (2017).

Google Scholar, Crossref, Indexed at

15. Yabroff, K. Robin, Katherine Reeder-Hayes, Jingxuan Zhao and Michael T. Halpern, et al. “Health Insurance Coverage Disruptions and Cancer Care and Outcomes: Systematic Review of Published Research.” J Natl Cancer Inst 112 (2020): 671-687.

Google Scholar, Crossref, Indexed at

16. Fogleman, Amanda J., Georgia S. Mueller and Wiley D. Jenkins. “Does where you live play an important role in cancer incidence in the U.S.?” Am J Cancer Res 5 (2015): 2314.       

Google Scholar, Indexed at

17. Klein, William M.P., Mary E. O’Connell, Michele H. Bloch and Susan M. Czajkowski, et al. “Behavioral research in cancer prevention and control: emerging challenges and opportunities.” J Natl Cancer Inst 114 (2022): 179-186.

Google Scholar, Crossref, Indexed at

18. Collins, Yvonne, Kevin Holcomb, Eloise Chapman-Davis and Dineo Khabele, et al. “Gynecologic cancer disparities: A report from the health disparities taskforce of the society of gynecologic oncology.” Gynecol Oncol 133 (2014): 353–361.

Google Scholar, Crossref, Indexed at

19. Jura, Magdalena and Leslie Kozak. “Obesity and related consequences to ageing.” Age (Dordr) 38 (2016): 1-18.

Google Scholar, Crossref, Indexed at

20. De Angelis, Cristina, Antonio Nardone, Francesco Garifalos and Claudia Pivonello, et al. “Smoke, alcohol and drug addiction and female fertility.” Reprod Biol Endocrinol 18 (2020): 1-26.

Google Scholar, Crossref, Indexed at

21. Sirovich, Brenda E and H. Gilbert Welch. “The frequency of Pap smear screening in the United States.” J Gen Intern Med 19 (2004): 243-250.

Google Scholar, Crossref, Indexed at

22. Diaz-Montes, Teresa P and Robert L. Giuntoli. “Volume-based care among young women diagnosed with uterine cancer.” ISRN Surg 2011 (2011).

Google Scholar, Crossref, Indexed at

23. Schramm, Katharina, Juliane Nees, Janine Hoffmann, Thomas Bruckner, et al. “Emergency consultations in obstetrics: Identification of decisive, contributing and associated factors.” Arch Gynecol Obstet 302 (2020): 821-828.

Google Scholar, Crossref, Indexed at

24. Ezendam, Nicole P.M., Kim A.H. Nicolaije, Dorry Boll, Marnix L.M. Lybeert, et al. “Health care use among endometrial cancer survivors: A study from profiles, a population-based survivorship registry.” Int J Gynecol Cancer 23 (2013): 1258-1265.

Google Scholar, Crossref, Indexed at

25. Rosenman, Robert, Vidhura Tennekoon and Laura G. Hill. “Measuring bias in self-reported data.” Int J Behav Healthc Res 2 (2011): 320.

Google Scholar, Crossref, Indexed at