Where does the money go to? Cost analysis of gynecological patients with a benign condition

The impact of common benign gynecological conditions on the life of women [ 1 ] and on the costs for the society is high [ 2 – 6 ]. Half of women suffer from a benign gynecological condition during their fertile life [ 7 ]. Nearly 80% of them see a doctor annually and one third spends one or more days in bed every year [ 7 ]. Especially physical, sexual and mental well-being is impaired [ 1 , 8 , 9 ]. Costs of benign gynecological conditions are high and similar to those of chronic diseases such as diabetes, Crohn’s disease, Parkinson’s disease, and rheumatoid arthritis [ 10 , 11 ]. Gynecological bleeding disorders, for example, account for annual costs of $13 billion [ 8 ] and endometriosis for $22 billion (0.7% of US annual health care expenditure) in the US [ 12 , 13 ]. Limited healthcare resources mandate that more focus is set on optimizing healthcare processes and on the cost-effectiveness and opportunity costs of treatment. As the ageing of the population and the desire for improved quality-of-life are associated with high healthcare costs [ 14 ], it is important to understand how the total costs build up to be able to cut unnecessary costs and to allocate resources wisely. This study provides new, original, short- and long-term data on the healthcare processes and costs of treatment of common benign gynecological conditions in a publicly funded healthcare setting. Our study also reports the indirect costs of treatment and the more seldom reported other out-of-hospital costs. Furthermore, the study is set to answer whether the distribution of costs differs depending on the treated condition or the hospital size and whether hospital treatment diminishes the need to seek care elsewhere and thus decreases costs and indirect societal costs. This knowledge enables better recognition of patient groups whose treatment protocols need further development.

Of the 1173 women invited to participate, 543 (46.3%) completed the baseline questionnaire. Of these, 425 (78.3%) answered also at six months and 397 (73.1%) at two years. The inclusion criteria (appropriate reason of treatment and having answered all questionnaires at all follow-ups) were met by 389 patients at six months and 311 patients at two years. Of them, 178 (57.2%) were treated in the university hospital and 133 (42.8%) in community hospitals. Patients were treated because of bleeding disorders (n = 94), uterine fibroids (n = 79), polyps (n = 68), dysmenorrhea/pelvic pain (n = 24), endometriosis (n = 23), uterine prolapse (n = 7) or some other benign uterine disorder (n = 16). The mean duration of the long-term follow-up was 24 months 11 days (SD±12days). The mean age of the patients at baseline was 49.6 years (SD±12.8, range 19 to 98). Of all patients, 167 (46.3%) underwent surgery in an operation theatre. Of university hospital patients 44.4% (n = 79) and of community hospital patients 48.9% (n = 65) were treated operatively (difference not statistically significant). Patients with uterine fibroids or a prolapse were most often treated operatively (70.9% and 71.4%, respectively) and patients with bleeding disorders, pelvic pain or some other benign gynecological condition least often (31.9%, 20.8%, and 6.3%, respectively). The only statistically significant difference in operative treatment between the hospitals was that polyp patients were more often treated in the operation theatre in the community hospitals compared to the university hospital (66.7% versus 36.8%) ( S1 Fig ). The total direct costs averaged 688.6€ (SD±1098.8 €) at six months and 2193.8€ (SD±2158.6€) at two years. Treatment of uterine fibroids was the most expensive intervention at six months. At two years, the costs were highest in the endometriosis and uterine fibroids groups. The treatment of other benign gynecological conditions was less expensive during the whole follow-up ( Fig 1 ). There was no statistically significant difference between the mean costs of treatment of the whole patient group in the university or community hospitals either at six months (704.7€ vs. 667.0€, respectively) or two years (2172.5€ vs. 2222.2€, respectively). More resources were used to treat endometriosis patients in the university hospitals compared to the community hospitals both at six months and two years (means 807.7€ vs. 190.0€ and 3958.4€ vs. 2421.3€, respectively). By contrast, the university hospitals spent less money on the treatment of polyps in by the two-year follow-up (means 1535.8€ vs. 2049.5€, respectively). However, these differences were not statistically significant (p = 0.19, p = 0.13 and p = 0.14, respectively) ( Fig 1 ). Nearly half of total direct costs (48.4%) was caused by surgery. Doctor’s appointments, inpatient/ward treatment, and outpatient procedures all accounted for similar proportions (13–17%) ( Table 1 ). Doctor’s appointments generated more costs in community hospitals than in university hospitals. Otherwise, the cost distribution in the different hospitals was similar. A larger proportion of the costs was caused by ward and operation theatre treatment in patients with fibroids, endometriosis and prolapse when compared to the other diagnosis groups. Imaging caused relatively more costs in patients with pelvic pain or endometriosis. Outpatient procedures consumed relatively more resources in polyp patients and doctor’s appointments relatively more resources in the treatment of pelvic pain as compared to the other patient groups ( Table 1 ). Mean direct costs outside of the hospitals (pharmaceuticals, healthcare services outside of the hospital) decreased in the whole study group significantly compared to baseline (450.2€ ±339.7) at both the six-month (309.0€ ±204.9) and two-year follow-ups (281.3€ ±421.9) (p<0.001). This decrease was seen in patients treated in university hospitals (482.5€ vs. 294.5€ and vs. 279.6€, p<0.001 and p<0.005), but not in those treated in community hospitals (407.1€ vs. 328.4€ and vs. 283.6€, p = 0.14 and p = 0.11). These direct costs were similar for patients treated conservatively or operatively. However, the productivity costs were temporarily, at six months, higher for patients treated operatively than for those treated conservatively (627.7€ vs 165.0€, p<0.005), but returned to low levels by two years (58.3€ vs. 117.7€). The highest mean direct costs outside of the hospital were at baseline in the endometriosis group (735.0€ vs. 225.4–570.2€ in the other groups) and at six months in the pelvic pain group (598.3€ vs. 221.8€ - 370.5€). At the two-year follow-up, patients with a bleeding disorder, prolapse or some other benign gynecological condition reported significantly lower direct out-of-hospital costs than at baseline (p<0.05) ( Fig 2 ). At baseline, endometriosis and pelvic pain patients (423.7€ and 462.0€, respectively) and at six months patients with fibroids, pelvic pain or prolapse (686.9€, 1064.0€ and 888.0€, respectively) had the highest mean productivity costs compared to the other patient groups. In the long-term, however, marked productivity costs were observed only in endometriosis patients (803.5€ vs. 0–89.3€ of patients in the other groups). A significant decline in these costs during follow-up was noted in patients with a bleeding disorder or pelvic pain ( Fig 3 ). Absence from work and ultrasound examinations performed by private gynecologists generated the highest costs outside of the hospital at baseline followed by appointments with general practitioners at primary health care centres or private clinics, LNG-IUD purchases, and appointments with private gynecologists. The burden of sick leave declined from the six-month high back to baseline levels by two years, however, still accounting for the highest proportion of costs outside of the hospital. Already by six months, costs due to private gynecologists’ appointments had, instead, diminished. Appointments with general practitioners at primary health care centres decreased throughout the whole follow-up, whereas costs due to appointments with general practitioners at private clinics and purchases of LNG-IUD persisted throughout the follow-up. Purchases of painkillers, progesterone derivatives, and pharmaceuticals diminishing blood loss decreased from baseline to two years. Cost distribution was otherwise similar in patients of the university hospital and community hospitals throughout the study but only the costs of blood loss diminishing medicines declined significantly in the community hospitals ( S2 Fig ). At baseline, patients with pelvic pain used private gynecologists’ services least frequently but visited general practitioners most regularly. At six months, the difference in general practitioners’ visits at primary health care centres prevailed, but there were no longer differences in costs caused by private clinic general practitioner visits or private gynecological services between the diagnosis groups. At two years, also the use of general practitioners’ services at primary health care centres had evened out. At two years, endometriosis patients used private gynecological services significantly more than patients with a bleeding disorder or pelvic pain ( S2 Fig ). At baseline, patients with endometriosis or pelvic pain purchased more analgesics than the other diagnosis groups. This tendency persisted throughout the follow-up. At six months, endometriosis patients used significantly more oral progesterone derivatives but fewer bought LNG-IUDs compared to patients with a bleeding disorder. Patients with fibroids used blood loss decreasing medicines significantly more often at the short-term follow-up than patients with polyps. At the long-term follow-up, endometriosis patients purchased more pharmaceuticals to diminish blood loss than patients with a bleeding disorder or a polyp ( S2 Fig ).

A majority of direct costs does not arise during the first months of treatment but accumulate with time showing the need for prolonged healthcare management and cost follow-up. Lower costs could be achieved by fewer repeated doctors’ appointments, fewer monitoring tests and by re-evaluating the need for hospitalization and length of ward treatment. Procedures that do not need to be done in the operation theatre should be done ambulatorily. Hospital treatment reduces the need to seek medical attention outside of the hospital and decreases out-of-hospital expenses.

A prospective observational cost study conducted in the obstetrics and gynecology departments of the hospitals of the Helsinki and Uusimaa Hospital District: the Helsinki University Hospital (comprising of three hospitals in the Helsinki metropolitan area) and four smaller community hospitals. All women treated for a benign gynecological condition in these hospitals between June 1 st , 2012 and August 31 st , 2013 were invited to participate. Patients with uterine fibroids or some other benign uterine neoplasms, endometriosis, adenomyosis, uterine prolapse, endometrial or cervical polyps, gynecological bleeding disorder, dysmenorrhea or lower abdominal/pelvic pain or some other benign uterine finding were included in the study. The patients were grouped according to the ICD-10 classification codes after clinician’s routine assessment. If a specific condition/pathology was found behind the presenting symptoms (for example a uterine fibroid or a uterine polyp causing abnormal uterine bleeding), the patient was categorized according to this primary diagnosis. If no apparent reason for the condition was found, the patient was grouped according to this idiopathic symptom (for example idiopathic menorrhagia in the bleeding disorder group). Patients were treated according to typical clinical practice in the Finnish publicly funded (taxation based) healthcare system. The clinical care process was tracked regarding doctor’s appointments, laboratory examinations, imaging, pathological examinations, outpatient clinic procedures, surgical operations, and ward/inpatient treatment in the different patient groups. The direct hospital costs were acquired from the clinical patient administration database and are reported at the current cost level of the follow-up time (from June 2012 –August 2013 (study intake) until June 2014 –August 2015 (last follow-up survey). These direct hospital costs were analysed both in short-term (six-month follow-up) and in long-term (two-year follow-up). The average exchange rate during the study period varied between 1EUR = 1.11–1.33USD. To take out-of-hospital costs into account, the patients were asked to fill in a questionnaire concerning the use of services outside of the hospital before treatment (before/when attending the first outpatient clinic appointment, baseline), and at the six-month and the two-year follow-up. Patients reported nurses’ and doctors’ appointments at primary health care centres or private clinics, laboratory examinations, imaging, treatment for the same condition in other hospitals or clinics, drug purchases (including pain killers (for example NSAIDs), blood loss diminishing medicines (for example tranexamic acid), progesterone derivates (hormonal products) and separately LNG-IUD (levonorgestrel-releasing intrauterine device)) and absence from work during a three-month time period before each follow-up. Healthcare unit costs, productivity costs in Finland [ 15 , 16 ] and pharmaceutical prices were incorporated into patient-reported use of services to obtain estimates of total costs to the patient, hospitals and society throughout the follow-up. Paired samples t-test was used to test the statistical significance of the change in costs during follow-up. The statistical significance of the differences in the mean costs between two unrelated groups was tested by independent samples t-test. Two-sided p-values < 0.05 were considered statistically significant. The data were analysed using SPSS for Windows statistical software version 22.0 (SPSS, Inc., Chicago, IL, USA). The Ethics Committee of the Helsinki University Central Hospital approved the study protocol (registration number 538/E0/02). The study was conducted in accordance with the 1964 Helsinki declaration and its later amendments. Informed written consent was obtained from all participants.

(DOCX) Click here for additional data file. (DOCX) Click here for additional data file. (XLS) Click here for additional data file.