Temporal and geospatial trends in male factor ... - Fertility and Sterility

Download PDF
4 downloads 15373 Views 848KB Size Report
Comment
Jun 12, 2014 - a Department of Urology, University of California, San Francisco, San Francisco, ... within the context of assisted reproductive technology (ART).
Temporal and geospatial trends in male factor infertility with assisted reproductive technology in the United States from 1999–2010 Anobel Y. Odisho, M.D., M.P.H.,a Ajay K. Nangia, M.B.B.S.,b Patricia P. Katz, Ph.D.,a,c and James F. Smith, M.D., M.S.a,d a Department of Urology, University of California, San Francisco, San Francisco, California; b Department of Urology, University of Kansas Medical Center, Kansas City, Kansas; c Department of Medicine and Institute for Health Policy Studies, University of California, San Francisco, San Francisco, California; and d Department of Obstetrics, Gynecology, and Reproductive Sciences and Institute for Health Policy, University of California, San Francisco, San Francisco, California

Objective: To estimate the prevalence of male factor infertility diagnosis within the context of assisted reproductive technology (ART) clinics and its geographic and temporal distribution from 1999–2010. Design: Population study based on patients presenting for care at ART centers. Setting: Clinics providing ART services. Patient(s): All male patients seeking infertility care at ART clinics. Intervention(s): Data were obtained from the Centers for Disease Control and Prevention, analyzed, geocoded, and mapped. Main Outcome Measure(s): Prevalence of male factor infertility diagnosis in a couple seeking infertility care. Result(s): Between 1999 and 2010, 1,057,402 cycles of ART using nonfrozen, nondonor eggs were performed, increasing from 62,809 cycles in 1999 to 99,289 cycles in 2010. Nationwide in ART clinics, the period prevalence of isolated male factor infertility was 17.1% and the prevalence of overall male factor infertility diagnoses was 34.6%. The highest prevalence was reported in New Mexico (56.4%) and lowest in Mississippi (24.2%). Conclusion(s): The prevalence of male factor infertility diagnosis varies significantly by time and space within the United States, whereas its overall prevalence has remained remarkably stable. This study provides the spatial Use your smartphone analytic framework for future research to explore factors associated with male factor infertility. to scan this QR code (Fertil SterilÒ 2014;102:469–75. Ó2014 by American Society for Reproductive Medicine.) and connect to the Key Words: Male factor infertility, prevalence, geospatial mapping, GIS, health services Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/odishoa-temporal-geospatial-male-infertility-art-us/

S

ignificant social and demographic shifts in the United States, combined with improved IVF techniques, have led to increased interest in and utilization of assisted reproductive technology (ART). Men and women continue to postpone marriage, with a current median age of first marriage at 25.8 years for women and 28.3 years

for men (1). The mean maternal age at first childbirth has increased, and the percentage of births to women more than 40 years has more than doubled, from 1.2%–2.8%, in the past 20 years (2–4). Although knowledge of female infertility prevalence allows clinicians and researchers to examine underlying causes for these changes, the

Received January 30, 2014; revised April 27, 2014; accepted May 8, 2014; published online June 12, 2014. A.Y.O. has nothing to disclose. A.K.N. has nothing to disclose. P.P.K. has nothing to disclose. J.F.S. has nothing to disclose. Reprint requests: James F. Smith, M.D., M.S., Department of Urology, University of California, San Francisco, San Francisco, California 94143-1695 (E-mail: [email protected]). Fertility and Sterility® Vol. 102, No. 2, August 2014 0015-0282/$36.00 Copyright ©2014 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2014.05.006 VOL. 102 NO. 2 / AUGUST 2014

discussion forum for this article now.*

* Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.

prevalence and distribution of male factor infertility, thought to be present in 40%–50% of infertile couples, remains poorly understood (5–7). Survey-based measures of the prevalence of male factor infertility have shown that 7.5% of men in the United States report seeking help for infertility, and of those that sought care, 18.1% reported clinician-diagnosed malerelated infertility (8). More than 1.1 million men sought fertility care in 2002 (9) and there were 131–172 physician visits per 100,000 insured men for infertility care between 1994 and 2006 (10, 11). National estimates, using data from the National Survey of Family 469

ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY Growth, suggest that the prevalence of male factor infertility was approximately 12% in 2002 (12). At present, however, no study has estimated the geographic distribution of a male factor infertility diagnosis in the United States at the national and regional levels or explored changes in male factor infertility diagnoses over time. The objective of our study was to characterize longitudinal and geographic trends in the diagnosis of male infertility within the context of ART clinics in the United States during a 12-year period and to identify regions or time periods of high or low levels of male factor infertility.

MATERIALS AND METHODS Data Preparation As a result of the Fertility Clinic Success Rate and Certification Act passed in 1992 (13), the data from all ART cycles performed in US fertility clinics are reported to the Centers for Disease Control and Prevention National ART Surveillance System. The National ART Surveillance System reports data for 95% of ART clinics. Annual infertility diagnoses data at the clinic level were obtained from the Centers for Disease Control and Prevention (14). Institutional Review Board approval was not required for this analysis of publicly available, deidentified, and aggregated data. Because clinic names were not standardized over time (the same clinic was reported over time using varying abbreviations, punctuation, and spellings), we used text clustering algorithms in Open Refine, a free open-source text, and data processing platform (15, 16). This was manually verified based on clinic names and their medical directors. Clinic addresses were obtained from the Centers for Disease Control and Prevention. Data and missing fields were updated with doubledata entry using Amazon Mechanical Turk (Amazon.com, Inc.) (17). Mechanical Turk allows for simple, repetitive human tasks to be assigned to a distributed workforce. Using this service allowed for rapid, high quality duplicate data entry for a low cost.

Geographic Analysis Geocoding, the process by which street addresses are converted into longitude and latitude coordinates, was performed using the R statistical computing environment (18) and the ggmap package (19) with the Google Maps API (20). Using the clinic longitude and latitude and the spatial package (21), clinics were overlaid on state and county maps obtained from the US Census TIGER/Line system (22) and Hospital Referral Region (HRR) maps obtained from the Dartmouth Atlas of Health Care (23). The HRRs are geographic areas developed by the Dartmouth Atlas of Health Care to represent markets for tertiary medical care delivery. The HRRs have been used in a number of settings to account for variations in healthcare access, utilization, and outcomes (24–26). This is a widely used health policy tool, more accurate than mapping purely along arbitrary political boundaries that do not reflect the distribution and utilization of healthcare resources.

Data Visualization Shaded matrix graphs, similar to gene microarray heat maps, were generated using R. They allowed for clear visual repre470

sentation of more than 600 data points and concisely displayed changes in male factor infertility prevalence over time. Choropleth maps, thematic maps with shading representing a calculated variable, were generated using QGIS, an open-source Geographic Information System platform (27).

Outcome Measure The prevalence of male factor infertility at ART clinics, defined as any abnormal semen parameter or sperm functional assay, was reported annually by each clinic, and for analysis was weighted by the number of nondonor, nonfrozen ART cycles performed by the clinic in that year. Clinics were the baseline unit of analysis, which were aggregated to state levels. Means were weighted based on proportional contribution of each clinic based on cycles performed in that state during the cumulative time period 1999–2010, as shown: X

ri 

cyclesi cycless

where i ¼ clinic; ri ¼ rate of male infertility as reported at each individual clinic; cyclesi ¼ number of IVF cycles performed at that clinic; and cycless ¼ cumulative number of IVF cycles performed in each state from 1999–2010. Isolated male infertility was defined as a diagnosis of male factor infertility without concomitant female factor infertility. Total male infertility was defined as any male factor infertility, either isolated or in the presence of female factor infertility. Data were aggregated to the state and HRR levels for analysis and mapping.

RESULTS The number of ART clinics in the United States increased from 367 in 1999 to 440 in 2010, representing a 19.9% increase (4.6% annual average). By 2010 only three states (Maine, Montana, and Wyoming) lacked an ART clinic. During this time period, 1,057,402 nondonor, nonfrozen cycles were performed, increasing from 62,809 in 1999 to 99,289 by 2010—a 58.1% increase (Table 1). The number of annual cycles peaked at 102,924 in 2008, with a slow decline through 2010. The mean annual prevalence of isolated male factor infertility as diagnosed at all ART clinics was 16.9%–17.5% in the years between 1999 and 2010, with an overall mean of 17.1%. Despite stability in the overall prevalence of male factor infertility, there was high variability at the state and HRR levels. Analysis by geographic region showed little significant variation in the diagnosis of isolated male factor infertility. Isolated male factor infertility was reported in 17.9% of men presenting to ART clinics in the Midwest, 16.8% in the Northeast, 17.1% in the South, and 16.9% in the West (SD 8.0%–9.8%). State-level analysis (Table 2) revealed high variability in isolated male factor infertility diagnoses. Utah (26.9%), Minnesota (25.4%), Wisconsin (24.6%), New Hampshire (24.2%), and Vermont (23.2%) had the highest proportion of isolated male factor infertility diagnosed at ART clinics, whereas Alabama (10.3%), Mississippi (11.3%), West Virginia (11.9%), Georgia (12.3%), and New Mexico (12.6%) had the VOL. 102 NO. 2 / AUGUST 2014

Fertility and Sterility®

TABLE 1

TABLE 2

Annual assisted reproductive technology (ART) data: clinics, cycles, and prevalence of male factor infertility diagnoses.

Cumulative prevalence of male factor infertility diagnosis by state, 1999–2010.

Nonfrozen, nondonor Isolated male factor Total male factor cycles Mean (SD)a % Mean (SD)b %

States

Cycles Isolated male Total male per 10k factor factor Cycles people % Mean (SD)a % Mean (SD)b

Mississippi Connecticut Maryland Arkansas Louisiana Illinois Colorado Indiana Nevada North Carolina Rhode Island Delaware Massachusetts Georgia New York South Carolina Pennsylvania New Hampshire California Washington, DC Arizona Maine Oregon Alaska Ohio North Dakota Washington Vermont Missouri New Jersey Kansas Iowa Texas Kentucky Idaho Virginia Oklahoma South Dakota Florida West Virginia Tennessee Nebraska Michigan Minnesota Alabama Hawaii Wisconsin Utah New Mexico

2,351 2,530 4,606 2,158 7,131 7,759 1,538 1,597 7,726 1,923 7,986 3,815 8,610 1,973 1,384 6,517 3,202 1,378 1,308 1,206 1,333 168 7,325 420 3,000 1,165 1,617 1,172 1,236 7,251 8,112 7,094 5,667 4,675 2,052 2,131 5,718 1,472 4,929 1,146 9,561 5,562 2,684 1,829 6,843 5,405 7,438 5,863 1,572

Year 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Total

Clinics (n) 367 380 381 388 396 410 419 424 427 434 438 440

62,809 71,250 76,729 81,599 86,418 89,467 92,022 93,638 100,377 102,924 100,880 99,289 1,057,402

17.52 (8.80) 16.93 (8.77) 17.07 (8.86) 17.06 (8.77) 17.17 (8.79) 17.00 (9.24) 16.87 (8.99) 16.91 (8.45) 17.05 (8.40) 16.98 (8.74) 17.43 (8.87) 17.41 (9.17) 17.12 (8.83)

34.07 (12.99) 33.62 (13.71) 34.43 (13.23) 34.79 (13.38) 34.32 (14.45) 34.54 (15.29) 34.75 (15.26) 34.48 (15.81) 34.94 (15.96) 34.56 (15.39) 34.89 (15.59) 35.08 (16.08) 34.56 (14.90)

a For each year, means arePweighted based on annual proportional of the number IVF cycles cyclesi at each clinic, as shown: ri  cycles a i ¼ clinic, each year; ri ¼ rate of male infertility as reported at each individual clinic; cyclesi ¼ number of IVF cycles performed at that clinic in each year; and cyclesa ¼ total number of IVF cycles performed nationwide in each year. b Isolated male factor infertility diagnosed at ART clinic as any semen analysis or sperm functional assay abnormality.

Odisho. Geospatial trends in male factor infertility diagnoses. Fertil Steril 2014.

lowest prevalence. Three of the five states with the highest number of cycles had prevalences close to the national mean (New York 16.5%, California 17.1%, Massachusetts 19.1%); however, two states were significantly lower (Illinois 14.2%, New Jersey 14.6%). When accounting for population, Washington, DC and Massachusetts had the highest per capita utilization of ART cycles. The annual total male factor infertility diagnoses (isolated male factor and male factor in presence of female factor) at ART clinics at the state level from 1999–2010 were relatively stable, ranging from 33.6%–35.1% (Fig. 1). However, significant geographic and temporal variation was observed. Temporally, there was a distinct dispersion in the reported prevalence of male factor infertility, despite stable means. The SD increased from 12.99 in 1999 to 16.08 by 2010. Although some states, typically those with the largest number of IVF cycles, had levels approximating the national average (New York, California, Massachusetts, and Illinois), several states demonstrated increasing (Wisconsin, Minnesota, Utah, Hawaii, and New Mexico) and decreasing levels of male factor infertility over time (Delaware, Rhode Island). In states with high levels of infertility diagnoses, the increases accelerated after 2004. To better understand the geographic variation, the cumulative prevalence of male factor infertility as diagnosed at ART clinics was mapped at the HRR level (Fig. 2). These data highlight wide variation in prevalence of infertility diagnoses and large areas lacking access to an IVF clinic. Only 179 of 306 (58.4%) of HRRs had an IVF clinic at some point between 1999 and 2010.

DISCUSSION Assessment of infertility prevalence is dependent on reporting in the form of population sample surveys or clinic-based meaVOL. 102 NO. 2 / AUGUST 2014

8.12 72.86 83.36 7.80 16.04 61.22 33.14 25.59 33.01 22.22 75.56 45.75 134.00 22.01 72.24 15.32 25.84 10.70 36.66 210.85 22.88 1.29 20.24 6.35 26.23 18.19 25.83 18.98 21.44 84.26 29.52 23.95 24.90 11.23 14.49 28.47 16.10 18.94 28.38 6.31 16.05 31.76 26.81 35.91 15.00 42.74 13.50 23.74 8.23

11.25 (5.53) 15.27 (5.73) 18.96 (8.86) 17.07 (9.18) 17.14 (7.20) 14.24 (7.42) 13.21 (6.75) 12.84 (8.85) 13.66 (5.77) 17.21 (7.58) 19.95 (4.60) 13.46 (5.74) 19.07 (6.36) 12.30 (6.69) 16.54 (8.98) 15.97 (7.17) 16.68 (8.85) 24.21 (5.35) 17.13 (7.65) 20.72 (12.17) 15.93 (9.00) 16.14 (4.24) 19.45 (6.35) 21.08 (5.21) 19.18 (8.71) 19.29 (9.44) 18.21 (9.64) 23.20 (9.38) 18.68 (12.37) 14.56 (8.27) 19.65 (7.14) 19.56 (9.74) 17.40 (10.64) 15.96 (8.61) 14.81 (7.29) 16.93 (8.30) 22.67 (5.24) 17.73 (2.67) 18.00 (8.24) 11.86 (5.55) 13.58 (9.41) 20.83 (4.80) 20.66 (9.55) 25.44 (7.09) 10.29 (6.85) 14.11 (9.03) 24.56 (11.16) 26.89 (6.45) 12.57 (3.82)

24.16 (6.96) 24.47 (7.60) 26.55 (12.04) 26.73 (10.61) 26.94 (11.60) 28.48 (12.48) 28.76 (15.70) 28.76 (14.42) 29.25 (8.84) 30.24 (13.48) 30.46 (4.58) 30.53 (8.67) 30.81 (11.27) 30.82 (11.41) 32.67 (16.22) 32.81 (12.76) 33.03 (14.13) 34.52 (5.21) 34.55 (12.71) 34.85 (18.10) 35.38 (16.42) 35.67 (4.95) 35.67 (6.63) 35.72 (5.80) 35.72 (12.01) 36.26 (6.39) 36.34 (13.83) 36.38 (9.47) 36.80 (16.93) 36.96 (15.05) 37.73 (9.85) 38.14 (11.63) 39.56 (17.34) 39.63 (14.00) 39.75 (5.30) 39.79 (12.45) 40.51 (7.04) 40.66 (9.30) 41.21 (14.87) 42.79 (19.87) 42.79 (12.69) 45.47 (13.75) 45.66 (13.69) 47.49 (13.72) 50.01 (17.34) 50.52 (16.63) 51.66 (20.56) 52.66 (9.77) 56.35 (15.50)

a For each year, means arePweighted based on annual proportional of the number IVF cycles cyclesi at each clinic, as shown: ri  cycles a i ¼ clinic, each year; ri ¼ rate of male infertility as reported at each individual clinic; cyclesi ¼ number of IVF cycles performed at that clinic in each year; and cyclesa ¼ total number of IVF cycles performed nationwide in each year. b Isolated male factor infertility diagnosed at ART clinic as any semen analysis or sperm functional assay abnormality.

Odisho. Geospatial trends in male factor infertility diagnoses. Fertil Steril 2014.

sures. The Urologic Diseases in America project reported a slow downtrend of physician office visits for male factor infertility diagnoses per 100,000 privately insured men from 395.5 in 2002 to 273.3 in 2006 (10, 11). The National Survey of Family Growth demonstrated that 7.5% of all respondents reported seeking help to have a child, with 471

ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY

FIGURE 1

Annual prevalence of isolated male factor infertility diagnosed at assisted reproductive technology (ART) clinics in the United States from 1999–2010. Odisho. Geospatial trends in male factor infertility diagnoses. Fertil Steril 2014.

18.1% of them reporting a clinician-diagnosed male factor infertility problem. The unweighted number of respondents seeking help was only 224 of 4,109 respondents, with 40 of that subset reporting an infertility diagnosis, which is a limitation of the sampling and does not allow for analysis of geographic variation (8). Nonetheless, based on these results, Anderson and colleagues (8) estimated 3.3–4.7 million men

(2.4%–3.4% of the 2002 male population) reported lifetime visits for male infertility and 787,000–1.5 million had visits in the year preceding the survey. One of the primary challenges in understanding the epidemiology of male factor infertility is the accurate measurement of its prevalence. Clinic-based measures are prone to ascertainment bias, as the clinics are not standardized to

FIGURE 2

Prevalence of cumulative male factor infertility diagnoses by hospital referral region, weighted by number of IVF cycles performed in each hospital referral region between 1999 and 2010. Odisho. Geospatial trends in male factor infertility diagnoses. Fertil Steril 2014.

472

VOL. 102 NO. 2 / AUGUST 2014

Fertility and Sterility® guidelines for diagnosis of infertility nor do all men experiencing infertility present for care. In the present study, we used clinic-level data that were aggregated in a weighted fashion to the state level. Although this dataset may not show the true prevalence of infertility, it is useful to identify discrepancies in the rates of diagnosis. In our large population-based sample representing the entire United States and encompassing more than 1 million cycles during a 12-year observation period, the overall prevalence of isolated male factor infertility diagnosis was 17.1%. This is the first report of nationwide clinic-based data for male factor infertility and the large geographically associated sample size allowed evaluation of specific spatial and temporal trends during a long period of time. We observed significant male factor infertility variation with time and geographic location. Important, the overall nationwide prevalence of both isolated and combined male factor infertility remained remarkably stable. However, increased dispersion indicates higher variability in reported rates, particularly after 2004. This raises questions regarding the underlying etiology, as there have been no previously reported temporal trends. This may be related to changes in overall male health pattern trends (obesity, cumulative environmental exposures, cancer), increasing male age when starting a family, or important changes in rates and patterns of diagnosis and access to care. In addition, geographic hotspots like the Central Midwest and Utah must be more carefully examined to better understand drivers of male factor infertility diagnosis. Assessing both spatial and temporal trends in such large datasets presents significant challenges. Although mapping the data show interesting cumulative trends, we used a shaded matrix display to show spatial and temporal trends, yet still highlighting individual state level data. This revealed many trends, such as a temporal increase in variation of infertility diagnoses and the tendency for states with a high prevalence of infertility diagnoses to increase at an accelerated pace. Although states with the highest number of cycles tended to have similar proportion of male factor infertility diagnoses (California, New York, Massachusetts), many high volume deviations in geographic clusters were evident. Prior spatial analysis of ART service distribution showed that the Northeast has the highest percentage of the male population in their reproductive years living within a 60-minute drive to an ART center (28). In these high volume states with similar infertility diagnosis prevalence, 90%–97% of the male reproductive population lives within 60 minutes of an ART center. Data from the Urologic Diseases in America show that utilization is highest in the Northeast, with 228 age-adjusted physician office visits per 100,000 insured men in 2006, compared with 111–129 in the remainder of the country (10, 11). Michigan, Wisconsin, and Minnesota, contributing >52,000 combined cycles, reported a much higher prevalence of isolated male factor infertility (21%–25%) and slightly lower rates of reproductive age men living near an ART center (70%–80%). This may suggest an inaccurate male factor infertility diagnosis in areas that have limited access to either VOL. 102 NO. 2 / AUGUST 2014

ART centers with specialized high volume semen analysis labs or providers. This is supported by the high variability of male factor infertility in the South, a region that has significantly lower access to ART centers (28). In addition to proximity, socioeconomic status may potentially affect rates of diagnosis, and it has been shown that lower socioeconomic status significantly limits the number and types of treatments chosen by patients (29). There are many factors known to be associated with male infertility such as obesity (30, 31), prostate and testicular cancer (32, 33), and environmental factors (34, 35). Although none of these associations has been explored spatially at the national level, these dataset and spatial analytic tools can be used to define the extent and variation of male factor infertility presenting to ART centers and to define possible changes and hotspots with time. All clinics and the associated data were geocoded to a precise latitude and longitude, which has never been done for male factor infertility. This will allow for future analyses integrating other publicly accessible geographic datasets such as environmental exposure data (local air and water quality, distance to hazardous waste facilities, traffic density), U.S. census data, and local socioeconomic data. In addition, despite overall stability in prevalence, the use of intracytoplasmic sperm injection (ICSI) nationwide increased significantly faster. Future analyses could explore this relationship in detail. This study is limited by the ascertainment bias inherent in assessing a clinic-based sample of patients. The prevalence of infertility is limited to men seeking medical care for infertility, and in our study it is limited to those couples presenting for ART services. It is known that cohorts presenting for infertility care have a significantly higher likelihood of being white, married, higher socioeconomic status, and are better educated and older, which limits our ability to generalize to the total US population. Our study likely under-represents the true prevalence of male infertility in the United States, as not all couples or men with infertility have access to, seek, or need ART care (e.g., men evaluated and treated for correctable male factor infertility by urologists, endocrinologists, or other health care providers). Our methodology, with the temporal and spatial analysis allowed us to see this trend or ‘‘hotspots’’ and highlights the useful nature of this tool, but is limited in that it identifies associations but not causal relationships. It cannot be extended to draw conclusions about individual patients. The best ART registry available at present is the federally mandated Centers for Disease Control and Prevention ART Surveillance Registry. However, this registry provides limited diagnostic detail, defining male factor as only an ‘‘abnormal semen analysis,’’ without any etiologic detail (36). This may lead to over-reporting and under-reporting of isolated and male factor infertility, but again only in ART centers. The registry does not capture patients with infertility who did not seek care at ART centers, limiting our ability to draw population-level conclusions. In addition, of those that seek care at an ART center, not all progress to IVF. Prior work has shown that 55% of couples from a multicenter ART cohort underwent IVF (37). Also, because the data are reported as 473

ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY cycles at the clinic level, we were unable to account for individual patients undergoing multiple cycles or crossing state lines to seek care. These deficiencies in male factor infertility data highlight the need for standardization and improved nationwide databases/registries that would allow better reporting of the true prevalence of male factor infertility, a better understanding of the epidemiology of male factor infertility, and exploration of relationships with other disease processes and environmental risk factors. Despite these limitations, these data have significant strengths. Visual representation of male factor infertility trends suggests a number of potential explanations such as disparities in access to male reproductive care and urologists, under-reporting or over-reporting of male factor infertility, toxic environmental exposures associated with infertility, or, intriguingly, may reflect other underlying men's health problems associated with infertility such as obesity or systemic disease. Important, this study demonstrates the feasibility of geospatial analysis in studying men's health, allowing future studies to examine relationships between any geographically tagged dataset (e.g., for environmental exposures) and male reproductive health. Men with the overall diagnosis of male factor infertility presenting for ART in the United States has remained stable during a 12-year period, although the prevalence shows significant temporal and geographic variation at the state level. The high prevalence and increase in male factor infertility diagnoses in certain states is concerning and may be multifactorial in etiology. Identifying hot spots of male factor infertility diagnoses serves as a means to study male reproductive health as a marker of overall men's health or access to care and allows for correlation with potential demographic, socioeconomic, access to care, and public health concerns such as environmental and lifestyle. This study provides the spatial analytic framework for future research to examine relationships between any geographic dataset and male factor infertility. In addition, results have important implications for health policy, both to focus limited resources on areas of highest need and to explore impacts of insurance coverage mandates.

REFERENCES 1.

2.

3. 4.

5. 6. 7.

474

Copen CE, Daniels K, Vespa J, Mosher WD. First marriages in the United States: data from the 2006–2010 National Survey of Family Growth. Natl Health Stat Report 2012:1–21. Mills M, Rindfuss RR, McDonald P, te Velde E, ESHRE Reproduction and Society Task Force. Why do people postpone parenthood? Reasons and social policy incentives. Hum Reprod Update 2011;17:848–60. Kochanek KD, Kirmeyer SE, Martin JA, Strobino DM, Guyer B. Annual summary of vital statistics: 2009. Pediatrics 2012;129:338–48. Martinez G, Daniels K, Chandra A. Fertility of men and women aged 15–44 years in the United States: National Survey of Family Growth, 2006–2010. Natl Health Stat Report 2012;51:1–28. Irvine DS. Epidemiology and aetiology of male infertility. Hum Reprod 1998; (13 Suppl 1):33–44. Mosher WD. Fecundity and infertility in the United States. Am J Public Health 1988;78:181–2. Comhaire F. Towards more objectivity in the management of male infertility. The need for a standardized approach. Int J Androl 1987;Suppl 7:1–53.

8.

Anderson JE, Farr SL, Jamieson DJ, Warner L, Macaluso M. Infertility services reported by men in the United States: national survey data. Fertil Steril 2009; 91:2466–70. 9. Chandra A, Stephen EH. Infertility service use among U.S. women: 1995 and 2002. Fertil Steril 2010;93:725–36. 10. Meacham RB, Joyce GF, Wise M, Kparker A, Niederberger C. Urologic Diseases in America Project. Male infertility. J Urol 2007;177:2058–66. 11. Litwin MS, Saigal CS, eds. Male reproductive health. In: Urologoc Diseases in America. Washington, DC: US Department of Health and Human Services, Public Health Service, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, 2012: 405–43. 12. Louis JF, Thoma ME, Sørensen DN, McLain AC, King RB, Sundaram R, et al. The prevalence of couple infertility in the United States from a male perspective: evidence from a nationally representative sample. Andrology 2013;1: 741–8. 13. Centers for Disease Control and Prevention (CDC), Department of Health and Human Services (HHS). Implementation of the Fertility Clinic Success Rate and Certification Act of 1992. Proposed model program for the certification of embryo laboratories. Federal Register; 1998. Available at: http://www.gpo.gov/fdsys/pkg/FR-1998-11-06/pdf/98-29374.pdf. Last accessed May 29, 2014. 14. Centers for Disease Control and Prevention. CDC-ART Reports—Assisted Reproductive Technology Reports. Available at: http://www.cdc.gov/art/AR TReports.htm. Last accessed May 29, 2014. 15. GitHub. OpenRefine. Available at: https://github.com/OpenRefine/OpenRe fine. Last accessed May 29, 2014. 16. GitHub. Clustering In Depth. Methods and theory behind the clustering functionality in Google Refine. Available at: https://github.com/OpenRe fine/OpenRefine/wiki/Clustering-In-Depth. Last accessed May 29, 2014. 17. Amazon Mechanical Turk. Available at: http://www.mturk.com. Last accessed May 29, 2014. 18. R Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2013. Available at:, http://www.R-project.org. Last accessed May 29, 2014. 19. Kahle D, Wickham H, eds. ggmap. Available at: http://cran.r-project.org/ web/packages/ggmap/. Last accessed May 29, 2014. 20. Google Developers. Google, Inc. Available at: https://developers.google. com/maps/. Last accessed May 29, 2014. 21. Venables WN, Ripley BD. Modern applied statistics with S. New York: Springer; 2002. 22. US Census Bureau. TIGER Products—Geography. Available at: http:// www.census.gov/geo/maps-data/data/tiger.html. Last accessed May 29, 2014. 23. The Dartmouth Institute for Health Policy and Clinical Practice. Dartmouth Atlas of Health Care. Available at: http://www.dartmouthatlas.org/. Last accessed May 29, 2014. 24. Gottlieb DJ, Zhou W, Song Y, Andrews KG, Skinner JS, Sutherland JM. Prices don't drive regional Medicare spending variations. Health Aff (Millwood) 2010;29:537–43. 25. Song Y, Skinner J, Bynum J, Sutherland J, Wennberg JE, Fisher ES. Regional variations in diagnostic practices. N Engl J Med 2010;363:45–53. 26. Epstein AM, Jha AK, Orav EJ. The relationship between hospital admission rates and rehospitalizations. N Engl J Med 2011;365:2287–95. 27. Quantum GIS Development Team. Quantum GIS Geographic Information System. Available at: http://qgis.osgeo.org. Last accessed May 29, 2014. 28. Nangia AK, Likosky DS, Wang D. Distribution of male infertility specialists in relation to the male population and assisted reproductive technology centers in the United States. Fertil Steril 2010;94:599–609. 29. Smith JF, Eisenberg ML, Glidden D, Millstein SG, Cedars M, Walsh TJ, et al. Socioeconomic disparities in the use and success of fertility treatments: analysis of data from a prospective cohort in the United States. Fertil Steril 2011; 96:95–101. 30. Palmer NO, Bakos HW, Fullston T, Lane M. Impact of obesity on male fertility, sperm function and molecular composition. Spermatogenesis 2012;2:253–63. 31. Kasturi SS, Tannir J, Brannigan RE. The metabolic syndrome and male infertility. J Androl 2008;29:251–9.

VOL. 102 NO. 2 / AUGUST 2014

Fertility and Sterility® 32.

33. 34.

Walsh TJ, Schembri M, Turek PJ, Chan JM, Carroll PR, Smith JF, et al. Increased risk of high-grade prostate cancer among infertile men. Cancer 2010;116:2140–7. Hotaling JM, Walsh TJ. Male infertility: a risk factor for testicular cancer. Nat Rev Urol 2009;6:550–6. Jurewicz J, Hanke W, Radwan M, Bonde JP. Environmental factors and semen quality. Int J Occup Med Environ Health 2009;22:305–29.

VOL. 102 NO. 2 / AUGUST 2014

35. 36. 37.

Oliva A. Contribution of environmental factors to the risk of male infertility. Hum Reprod 2001;16:1768–76. Centers for Disease Control and Prevention. National ART Surveillance System. Available at: http://www.cdc.gov/art/NASS.htm. Last accessed May 29, 2014. Wu AK, Odisho AY, Washington SL, Katz PP, Smith JF. Out-of-pocket fertility patient expense: data from a multicenter prospective infertility cohort. J Urol 2014;191:427–32.

475