Article Text
PDF
Abstract
Objective To determine if the introduction of value-based healthcare (VBHC) in fertility care can help to create realistic expectations in patients resulting in increased patient value, by demonstrating the relevance of defining outcome measures that truly matter to subfertile patients.
Design Retrospective cohort study.
Setting Tertiary fertility centre.
Results Time to pregnancy (TTP) and ongoing pregnancy rate (OPR), as a proxy for the live birth rate, for the full cycle of fertility care, regardless of which and how many treatment cycles performed, were identified as the most relevant medical outcome measures. Outcome measures were incorporated into a digital dashboard by using anonymised and validated patient data from the electronic patient file. We were able to present the TTP and OPR for the population as a whole as well as stratified for age, diagnosis, gravidity and type of gamete source used thereby resulting in a virtual ‘patient like me’ resembling the individual patient in the consultation room.
Conclusion We have shown that, by applying VBHC principles, relevant outcome measures can be generated and stratified for different patient characteristics, in order to develop a virtual ‘patient like me’. This virtual ‘patient like me’ can be used in the consulting room in the form of a digital dashboard, attributing to create realistic patient expectations. This facilitates healthcare providers and patients in shared decision-making.
- subfertility
- patient reported outcome measures
- decision making
- reproductive medicine
Data availability statement
The data underlying this article will be shared on reasonable request to the corresponding author and after approval of a written proposal taking into account Dutch legislation and Good Practice guidelines.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
The time to pregnancy and the ongoing pregnancy rate for the full cycle of fertility care serve as a means for shared decision-making.
These outcome measures can be stratified by different patient characteristics resulting in a virtual ‘patient like me’, to be presented in the consultation room.
As a result, realistic expectations can be created in patients and healthcare staff.
Results are based on a retrospective analysis in a single centre and need to be validated in other fertility clinics.
Introduction
Value-based healthcare (VBHC) was first introduced by Porter1 as a new method for providing medical care by increasing patient value through measuring truly, for the patient, relevant outcome measures and knowing the costs to achieve these outcomes. This method is currently being embraced and implemented into medical practice worldwide.2 Many clinics are introducing this concept and International Consortium for Health Outcome Measurement (ICHOM) sets with relevant outcomes of care are created for a multitude of medical conditions.3
This strategy of improving patient value by defining and enhancing relevant outcome measures, has not yet been introduced in fertility care. Success rates in fertility care are often presented as (ongoing) pregnancy rates and live birth rates per treatment cycle. These outcome measures however, do not represent the efficacy of the full cycle of care, including fertility assessment and all treatment cycles performed. The full cycle of fertility care can be defined starting with a referral to a fertility clinic and ending with discharge after achieving an ongoing pregnancy (ending in a live birth) or after the fertility team and/or the patient decides to discontinue fertility treatment.
VBHC does not only take medical outcome measures into account, it considers the patient-reported outcome measures (PROMs) and patient-reported experience measures (PREMs) as well. Couples with subfertility are known to be burdened with a long process of fertility assessment and treatment, often being accompanied by major psychosocial implications.4 The perception and fear of being subfertile can cause depression, anxiety and chronic stress, the severity of which is comparable to other serious medical conditions.5 During treatment, the ongoing cycle of hope, expectation and disappointment due to the limited pregnancy rates contributes to these feelings and can induce a feeling of loss and mourning.4 6 Addressing the social and emotional impact of subfertility is an important aspect in the treatment process.7 Gathering data using PROMs can help to identify which patients are in need for psychological evaluation and guidance.
By defining, measuring and improving outcome measures that truly matter to subfertile couples, VBHC can improve the quality of the fertility care pathway. More importantly, it generates realistic expectations in patients, which in turn will serve as a basis for shared decision-making. Furthermore, by sharing standardised, clear and objective information about individual processes in fertility treatment and outcomes, all fertility care facilities will accomplish transparency. This will allow the comparison of (inter)national results more objectively, which could improve fertility care in general and patient satisfaction in particular.
This paper describes how to introduce VBHC in fertility care by defining outcome measures that are truly relevant to the patient for the full cycle of fertility care, as a first step. By incorporating this information, generated from the clinic’s own anonymised electronic patient files, into a digital dashboard, it can be used by the clinicians counselling the subfertile couples in shared decision-making.
Material and methods
Patient and public involvement
Patients were part of a multidisciplinary integrated practice unit (IPU) at the Isala Fertility Center in Zwolle, The Netherlands. The IPU defined medical outcome measures, PROMs and PREMs that are truly relevant to subfertile couples, since an ICHOM set is not available for fertility care. The proposed outcome measures were tested in a questionnaire, which was sent to 20 randomly selected patients of our centre (response rate 70%). These responses were then validated in another questionnaire, which was sent to 40 randomly selected patients (response rate 67.5%), in order to conclude whether these medical outcome measures, PROMs and PREMs were actually considered important by many and if so, to what extent, using a Likert 5-point scale.8
Database validation
Data of all patients who were referred to the Isala Fertility Center between 1 January 2015 and 31 December 2019 (n=7255) were extracted from the electronic patient file. Data from male patients, data from patients with an unknown first clinic visit (FCV) date and data from patients who received part of their treatment in an affiliated fertility centre were excluded. This resulted in the inclusion of anonymised data of 4768 patients. To validate the robustness of these exported data, a sample size test was performed, by manually comparing data of 60 randomly selected files (using the online research randomiser www.randomizer.org) to their original source, that is, the electronic patient file.9
After validating the sample size for robustness, the entire database was checked for the percentage of missing data in order to determine the reliability of each individual parameter. The following individual parameters were assessed: patient identification number, sex, FCV date, age at FCV, diagnosis, gravidity (as a proxy for primary or secondary subfertility), gamete source, live birth, ongoing pregnancy, first day of the last menstrual period (LMP) of the cycle resulting in a pregnancy and time to pregnancy (TTP) which was defined as LMP minus FCV.
The parameters gravidity and diagnosis were subsequently studied in more detail, since a considerable number of files had been left blank for these parameters. An additional set of 60 files was randomly selected for both parameters (using the online research randomiser mentioned above) and manually compared with the original source.
Data analysis
The median TTP and ongoing pregnancy rate (OPR) were calculated for the entire population and were stratified by age at FCV, diagnosis, gravidity and gamete source parameters. Age categories used in previously published fertility studies were applied by categorising age (in years) at FCV as younger than 30, age between 30 and 34, age between 35 and 39 and age 40 or above.10 11 The diagnosis parameter was subdivided into the following categories: male factor, ovulatory factor, tubal factor, combination of factors (if two or more causes of subfertility were identified), unexplained and other (including cervical factor, carrier of genetic disease, endometriosis, fertility preservation, gamete/embryo donation, habitual abortion, sexual dysfunction and uterine factor). Gravidity subgroups were defined as primary (G0) and secondary subfertility (≥G1). The parameter gamete source was subdivided into the categories partner, sperm donor, and egg donor.
A test for normal distribution was performed using MedCalc (MedCalc Statistical Software V.19.0.5; MedCalc Software bvba, Ostend, Belgium). The parameter TTP was not normally distributed; therefore the median value and IQRs are reported.
Results
Defining outcome measures
As demonstrated below, the parameter live birth could be insufficiently reliable and extracted from the electronic patient files. Therefore, as a proxy, the OPR was selected. After analysing the questionnaires and deliberation by the IPU, it was established that TTP and OPR (as a proxy for live birth rate) for the full cycle of fertility care were considered most relevant medical outcome measures for subfertile couples.
Database robustness validation
The 60 randomly selected files showed that the data of the sample size matched the electronic patient file for all relevant parameters, except for three.
Sixteen documented FCVs did not match the actual FCV according to the electronic patient file (26.7%). The documented FCV is automatically generated on the date a new patient file is created by the treating physician. This is not necessarily performed on the actual date a patient visits the fertility clinic for the first time. Based on FCV, the system automatically generates a number for TTP when an ongoing pregnancy is documented (LMP minus FCV). A margin of error of this non-normally distributed parameter was therefore calculated. It was found that the median difference in days between the registered FCV and the actual FCV was 5.5, which was considered a clinically irrelevant margin of error.
During the validation process, it was found that in a significant number of files, the parameter gravidity had not been registered (table 1). In depth analysis of the complete electronic patient file of 60 additional patients showed that when gravidity was not registered, a previous pregnancy could not be demonstrated and therefore these patients could be considered nulligravida. This group was therefore added to the group of primary subfertile patients.
Degree of documentation of different parameters in the electronic patient file
Within the parameter gamete source, the subcategory egg donor has not been included in the results, due to the low number of patients. For the parameter diagnosis, in 47% of files, data were missing in the intended field, but a clear diagnosis was found in another part of the electronic patient file.
Drop-outs (subfertile couples discontinuing treatment without medical advice to do so) were not excluded from the calculations in order to obtain data resembling daily practice.
Database reliability validation
In order to conclude whether or not the parameters were reliable, the degree of documentation was checked for every parameter for all 4768 files (see table 1). The level of accuracy varied for parameters which had to be added manually in the electronic patient file. TTP was generated in 99% of cases where an ongoing pregnancy had been established. LMP was documented in 100% of cases where a fertility treatment had actually started.
The parameter live birth was documented after a previously confirmed ongoing pregnancy. In 32% of cases this parameter was entered as unknown. In 27% of these cases, the due date had not been reached yet at the time of analysis. In the other 73% of the cases the last consultation had been more than a year prior to analysis and patients should have given birth. This can be explained by patients being lost to follow-up after referral to their local midwife or medical centre. The parameter live birth was therefore considered insufficiently reliable and prone to bias. Hence, as a proxy for live birth, ongoing pregnancy beyond 12 weeks of gestation was used.
TTP and OPR
In our database, an OPR of 39.2% (95% CI: 37.81% to 40.59%) with a median TTP of 237 days (IQR: 108–463) was found. Stratification of these results was performed for age, diagnosis, gravidity and type of gamete source used (table 2). The practical implementation of this will be demonstrated by two fictional cases.
OPR and TTP stratified by different parameters
Case 1
A 28-year-old patient visits our clinic due to primary subfertility based on mild to severe oligoasthenozoospermia in the male partner.
When these patient characteristics were analysed separately, the following results were generated from our database: median TTP and the OPR of patients younger than 30 with subfertility due to any cause is 236.5 days (IQR: 102–463) and 46.7% (95% CI: 44.4% to 49.0%), respectively. Primary subfertility itself, without taking any other factors into account, has a median TTP of 245 days (IQR: 115–464.5) and an OPR of 40.2% (95% CI: 38.6% to 41.9%). For patients of any age with subfertility due to a male factor, a median TTP is 256 days (IQR: 131–454.8) with an OPR of 48.3% (95% CI: 44.9% to 51.7%). Lastly, if gametes of the partner of the patient were used in any fertility care pathway, the median TTP was 214 days (IQR: 99.5–406), with an OPR of 43.6% (95% CI: 41.9% to 45.2%).
Based on individual patient characteristics, results are diverse, with an OPR ranging from 40.2 to 48.3% and a median TTP ranging from 214 to 376.5 days.
When all the available parameters were combined, this particular patient is comparable to 249 patients in our database, showing a median TTP of 221.5 days (IQR: 187.9–269.8) and an OPR of 54.6% (95% CI: 48.4% to 60.9%) for the full cycle of care (figure 1).
Box-Whisker plot of TTP of the first fictional case, a 28-year-old patient with a primary subfertility and mild to severe oligozoospermia and/or asthenozoospermia in her male partner. The central box represents the values from the 25th to 75th percentile (Q2 and Q3), the middle horizontal line represents the median and the vertical line extends from the minimum to the maximum value (Q1–Q4) excluding outliers. TTP, time to pregnancy.
Case 2
A 39-year-old patient visits our clinic due to secondary subfertility based on a bilateral tubal factor and a partner with normozoospermia.
When these patient characteristics were analysed separately, the following results were generated. The median TTP and the OPR of patients between 35 and 39 years with subfertility due to any cause is 229 days (IQR: 104.8–430.3) and 29.4% (95% CI: 26.6% to 32.2%), respectively. Secondary subfertility itself, without taking any other factors into account, has a median TTP of 214 days (IQR: 82.5–461) and an OPR of 36.8% (95% CI: 34.3% to 39.3%). For patients of any age with subfertility due to a tubal factor, the median TTP is 329.5 days (IQR: 136–538) with an OPR of 37.7% (95% CI: 29.7% to 45.6%). Lastly, if gametes of the partner were used in any fertility care pathway, the median TTP was 214 days (IQR: 99.5–406), with an OPR of 43.6% (95% CI: 41.9% to 45.2%).
Based on individual patient characteristics, results are diverse, with an OPR ranging from 29.4% to 43.6% and a median TTP ranging from 214 to 329 days. When all available parameters were combined, this particular case is comparable to data of 20 patients in our database, showing a median TTP of 172 days (IQR 56.8–616.8) and an OPR of 36.8% (95% CI: 13.0% to 60.7%) for the full cycle of care (figure 2).
Box-Whisker plot of TTP of the second fictional case, a 39-year-old patient with a secondary subfertility based on a tubal factor and normozoospermia in her male partner. The central box represents the values from the 25th to 75th percentile (Q2 and Q3), the middle horizontal line represents the median and the vertical line extends from the minimum to the maximum value (Q1–Q4). TTP, time to pregnancy.
Discussion
In this study, we introduced the principles of VBHC in fertility care, in order to improve the quality of fertility care. As a first step, we focused on medical outcomes that are considered most relevant by subfertile patients. TTP and live birth/OPR for the full cycle of care are considered the most relevant outcome measures by our patients. We explored the use of these relevant outcome measures with the aim to generate accurate median TTP and OPRs for the full cycle of fertility care with the possibility to stratify these outcomes by different patient characteristics. This enables healthcare staff to apply the concepts of VBHC in the consulting room and to combine them with shared decision-making.
Shared decision-making is based on choice, option and decision talk and was eloquently described by Elwyn et al.12 In medical practice it is evident that correct and high-quality information is necessary to guide decision-making between various treatment options. The two abovementioned cases support the proof of concept that this also applies to the field of reproductive medicine. If TTP and OPRs are interpreted based on a single fertility treatment or a single patient parameter, it creates a non-representative perspective for the individual patient. We have shown that it is possible to generate more realistic perspectives if more relevant patient parameters are combined, thus increasingly matching the individual patient. This enables the patient, being guided by the healthcare provider, to generate realistic expectations which will support the shared decision-making process for the full patient journey in fertility care.
The data and results used are applicable to our specific patient population. If applied in other clinics, results might differ and therefore other centres should calculate their own medical outcomes. However, coming to a standardised outcome set will create the possibility for internal and external learning cycles. Over time, the database and the extracted results will become even more reliable, as data are added on a daily basis.
All parameters should be documented accurately by the clinicians and laboratory staff to be able to create a reliable database. As explained in the Results section, the parameter live birth was not a good representation of the actual live birth rate in our database. The proxy ongoing pregnancy was therefore used. It must be taken into account that perinatal death (fetal and neonatal death from 22 weeks of gestation up to and including 28 days after birth) occurs in 7.9‰ of pregnancies,13 meaning that the actual live birth rate will be slightly lower than the numbers reported in this paper. However, as Braakhekke et al14 showed, the OPR might reflect the effects of fertility treatment more accurately compared with live birth rate.
Perspectives, limitations and future directions
Using our validated database, we have shown that a digital dashboard can be developed to generate reliable estimates for median TTP and OPRs for the full cycle of care, for specific patient characteristics, with the possibility to stratify for different parameters. In future, the parameters body mass index, duration of subfertility and other predictive patient characteristics will be added to this application. Furthermore, the full cycle of care could be extended to the evaluation and referral of, for instance, a general practitioner and it could be the starting point to come to an internationally recognised standard on how to report and improve outcomes that are truly relevant for patients.
As mentioned before, the data of this study rely solely on the accurate documentation of manually documented parameters in the electronic patient file. A prospective study with clear instructions beforehand may further improve this. Moreover, a better and infallible registration method must be implemented to improve documentation of actual live births, in order to compute an accurate live birth rate. Future Information Technology (IT) solutions might catalyse the amount and accuracy of the parameters.
As a next step, we will focus on data concerning quality of life and psychosocial well-being as measured by PROMs for the full cycle of fertility care. Combined with the medical outcome measures as reported in this paper, this will generate a detailed and representative perspective for subfertile couples and their full fertility care pathway. Hereafter, a prospective analysis needs to be carried out to investigate our approach, including an evaluation of the economic advantages of a VBHC approach in fertility care. In the traditional pay-per-treatment strategies, healthcare providers might be incentivised to increase the number of fertility treatments and thereby costs without transparency whether this results in an increase truly for the patient-relevant outcome measures. Finally, our results need to be validated in other (international) fertility clinics.
Conclusion
This is the first study to show that, by applying VBHC principles, an accurate median TTP and an OPR can be generated for the full cycle of fertility care with the possibility to stratify for different patient characteristics. Based on our validated database, a digital dashboard is being developed to use in the consulting room of the fertility healthcare provider. We are convinced that it will contribute to a more transparent dialogue between doctor and patient, resulting in realistic expectations and shared decision-making. In future, such a fertility care dashboard may be used to compare full cycle of fertility care results between different centres. Future analyses are needed to focus on PROMs, to evaluate the economic advantages of a VBHC approach and to investigate this approach in other (international) fertility clinics.
Data availability statement
The data underlying this article will be shared on reasonable request to the corresponding author and after approval of a written proposal taking into account Dutch legislation and Good Practice guidelines.
Ethics statements
Patient consent for publication
Footnotes
Contributors MB and MHJMC designed the study. MB, MHJMC and JV prepared the manuscript. MB performed the validation and analysis of the data. CCAWvB and MHJMC defined the medical outcome measures and query to retrieve the data from the electronic patient file. WKHK, BJC and GT were responsible for the data and registration in the electronic patient file. All authors commented on and revised the manuscript and approved the final version. MHJMC as guarantor accepts full responsibility for the finished work and the conduct of the study, had access to the data, and controlled the decision to publish.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting or dissemination plans of this research. Refer to the Methods section for further details.
Provenance and peer review Not commissioned; externally peer reviewed.