The use of a new logistic regression model for predicting the outcome of pregnancies of unknown location

Diagnostic accuracy validation study of the M6 model without initial serum progesterone (M6NP) in triage of pregnancy of unknown location

OBJECTIVES

The M6 prediction model stratifies the risk of development of ectopic pregnancy (EP) for women with pregnancy of unknown location (PUL) into low risk or high risk, using human chorionic gonadotrophin (hCG) and progesterone levels at the initial visit to a gynaecological emergency room and hCG level at 48 h. This study evaluated a second model, the M6NP model, which does not include the progesterone level at the initial visit. The main aim of this study was to validate the diagnostic accuracy of the M6NP model in a population of French women.

STUDY DESIGN

Between January and December 2021, all women with an hCG measurement from the gynaecological emergency department of a teaching hospital were screened for inclusion in this study. Women with a pregnancy location determined before or at the second visit were excluded. The diagnostic test was based on logistic regression of the M6NP model, with classification into two groups: high risk of EP (≥5%) and low risk of EP (<5%). The reference test was the final outcome based on clinical, biological and sonographic results: failed PUL (FPUL), intrauterine pregnancy (IUP) or EP. Diagnostic performance for risk prediction of EP, and also IUP and FPUL, was calculated.

RESULTS

In total, 759 women with possible PUL were identified. After screening, 341 women with PUL were included in the main analysis. Of these, 186 (54.5%) were classified as low risk, including three (1.6%) with a final outcome of EP. The remaining 155 women with PUL were classified as high risk, of whom 60 (38.7%), 66 (42.8%) and 29 (18.7%) had a final outcome of FPUL, IUP and EP, respectively. Of the 32 women with PUL with a final outcome of EP, 29 (90.6%) were classified as high risk and three (9.4%) were classified as low risk. Therefore, the performance of the M6NP model to predict EP had a negative predictive value of 98.4%, a positive predictive value of 18.7%, sensitivity of 90.6% and specificity of 59.2%. If the prediction model had been used, it is estimated that 4.5 visits per patient could have been prevented.

CONCLUSION

The M6NP model could be used safely in the French population for risk stratification of PUL. Its use in clinical practice should result in a substantial reduction in the number of visits to a gynaecological emergency room.

Multiplexed serum biomarkers to discriminate nonviable and ectopic pregnancy

OBJECTIVE

To evaluate combinations of candidate biomarkers to develop a multiplexed prediction model for identifying the viability and location of an early pregnancy. In this study, we assessed 24 biomarkers with multiple machine learning-based methodologies to assess if multiplexed biomarkers may improve the diagnosis of normal and abnormal early pregnancies.

DESIGN

A nested case-control design evaluated the predictive ability and discrimination of biomarkers in patients at risk of early pregnancy failure in the first trimester to classify viability and location.

SETTING

Three university hospitals.

PATIENTS

A total of 218 individuals with pain and/or bleeding in early pregnancy: 75 had an ongoing intrauterine gestation; 68 had ectopic pregnancies (EPs); and 75 had miscarriages.

INTERVENTIONS

Serum levels of 24 biomarkers were assessed in the same patients. Multiple machine learning-based methodologies to evaluate combinations of these top candidates to develop a multiplexed prediction model for the identification of a nonviable pregnancy (ongoing intrauterine pregnancy vs. miscarriage or EP) and an EP (EP vs. ongoing intrauterine pregnancy or miscarriage).

MAIN OUTCOME MEASURES

The predicted classification using each model was compared with the actual diagnosis, and sensitivity, specificity, positive predictive value, negative predictive value, conclusive classification, and accuracy were calculated.

RESULTS

Models using classification regression tree analysis using 3 (pregnancy-specific beta-1-glycoprotein 3 [PSG3], chorionic gonadotropin-alpha subunit, and pregnancy-associated plasma protein-A) biomarkers were able to predict a maximum sensitivity of 93.3% and a maximum specificity of 98.6%. The model with the highest accuracy was 97.4% (with 70.2% receiving classification). Models using an overlapping group of 3 (soluble fms-like tyrosine kinase-1, PSG3, and tissue factor pathway inhibitor 2) biomarkers achieved a maximum sensitivity of 98.5% and a maximum specificity of 95.3%. The model with the highest accuracy was 94.4% (with 65.6% receiving classification). When the models were used simultaneously, the conclusive classification increased to 72.7% with an accuracy of 95.9%. The predictive ability of the biomarkers in the random forest produced similar test characteristics when using 11 predictive biomarkers.

CONCLUSION

We have demonstrated a pool of biomarkers from divergent biological pathways that can be used to classify individuals with potential early pregnancy loss. The biomarkers choriogonadotropin alpha, pregnancy-associated plasma protein-A, and PSG3 can be used to predict viability, and soluble fms-like tyrosine kinase-1, tissue factor pathway inhibitor 2, and PSG3 can be used to predict pregnancy location.

Follow-up and outcomes of patients with a pregnancy of unknown location: A comparison of two prediction models

BACKGROUND

The time period while delineating the final diagnosis following presentation with a pregnancy of unknown location (PUL) can be an anxious time, as well as being time and resource intensive. Prediction models have been utilised in order to tailor counselling, frame expectations and plan care.

OBJECTIVES

We aimed to review diagnoses of PUL in our population and assess the value of two prediction models.

STUDY DESIGN

We reviewed all 394 PUL diagnoses over a three year period in a tertiary level maternity hospital. We then retrospectively applied the M1 and M6NP models to assess their accuracy when compared to the final diagnosis.

RESULTS

PUL comprises of 2.9% (394/13401) of attendances in our unit, requiring 752 scans and 1613 separate blood tests. Just under one in ten women (9.9%, n = 39) presenting with a PUL had a viable pregnancy at discharge, however of the remainder, only 18.0% (n = 83) required medical or surgical treatment for a PUL. The M1 model was more successful at predicting an ectopic pregnancy than the M6NP, with the latter over-predicting viable pregnancies (33.4%, n = 77).

CONCLUSIONS

We demonstrate that the management of women with a PUL could be stratified through the application of outcome prediction models, having positive results for framing expectations and potentially reducing this resource-intensive diagnosis.

Predictors of Successful Medical Management With Methotrexate in Unruptured Tubal Ectopic Pregnancy

Introduction Medical treatment with methotrexate (MTX) is a safe and effective alternative to surgery in carefully selected cases of ectopic pregnancies diagnosed early prior to rupture. Aim To determine the optimal pre-treatment levels of beta human chorionic gonadotropin (𝛽-hCG) and its changing trends most likely to have a successful outcome with medical management. Material and methods A prospective observational study was conducted in a tertiary teaching hospital from December 2018 to May 2021. "Single-dose" MTX regime was used for medical management of ectopic pregnancy in patients fulfilling the selection criteria. The 𝛽-hCG levels were recorded at baseline and on day 4 and day 7 of MTX injection. Thereafter, at weekly intervals till complete resolution or surgical intervention due to failure of medical management. In addition, receiver operating characteristic (ROC) curve analysis for a pre-treatment 𝛽-hCG cut-off value and changing trends in post-treatment 𝛽- hCG levels most likely to have a successful outcome with MTX treatment were determined. Results Fifty patients fulfilling the inclusion criteria were included in the study, with successful medical management in 33 (66%). The mean pre-treatment 𝛽-hCG levels in women with successful medical management were 3270.97 (+/- 901) compared to 5249.17 (+/-808.02) for those with treatment failure (p=0.00001). The mean 𝛽-hCG level in the failed treatment group was significantly higher on day 4 than the pre-treatment levels (6742.56 +/- 572 vs. 5249.17+/- 808.02; p<0.05). Inadequate reduction of 𝛽-hCG level on day 7 (<15% of day 4 levels) requiring repeat dosage of MTX was more likely to have an unsuccessful outcome (p=0.00001). The area under curve (AUC) value of 0.905 (95% CI: 0.814-0.996) depicted that pre-treatment 𝛽-hCG level of 4000 mIU/ml taken as the cut-off value was able to predict levels ≤4000 mIU/ml had a greater likelihood of successful outcome with MTX, having a sensitivity of 84.5%, specificity of 83.3%, positive predictive value (PPV) of 90.3%, and negative predictive value (NPV) of 75% (p< 0.05). Demographic variables or previous clinical history, considered risk factors for ectopic pregnancy, did not affect the outcome of medical management in this study. Conclusion Medical management of ectopic pregnancy is a viable first-line treatment option in carefully selected patients. In this study, the most important predictors for the successful outcome of medical management were the pre-treatment β-hCG levels and their fall on day 4 and day 7 after MTX therapy.

Predictive analytical model for ectopic pregnancy diagnosis: Statistics vs. machine learning

Objective

Ectopic pregnancy (EP) is well known for its critical maternal outcome. Early detection could make the difference between life and death in pregnancy. Our aim was to make a prompt diagnosis before the rupture occur. Thus, the predictive analytical models using both conventional statistics and machine learning (ML) methods were studied.

Materials and methods

A retrospective cohort study was conducted on 407 pregnancies with unknown location (PULs): 306 PULs for internal validation and 101 PULs for external validation, randomized with a nested cross-validation technique. Using a set of 22 study features based on clinical factors, serum marker and ultrasound findings from electronic medical records, analyzing with neural networks (NNs), decision tree (DT), support vector machines (SVMs), and a statistical logistic regression (LR). Diagnostic performances were compared with the area under the curve (ROC-AUC), including sensitivity and specificity for decisional use.

Results

Comparing model performance (internal validation) to predict EP, LR ranked first, with a mean ROC-AUC ± SD of 0.879 ± 0.010. In testing data (external validation), NNs ranked first, followed closely by LR, SVMs, and DT with average ROC-AUC ± SD of 0.898 ± 0.027, 0.896 ± 0.034, 0.882 ± 0.029, and 0.856 ± 0.033, respectively. For clinical aid, we report sensitivity of mean ± SD in LR: 90.20% ± 3.49%; SVM: 89.79% ± 3.66%; DT: 89.22% ± 4.53%; and NNs: 86.92% ± 3.24%, consecutively. However, specificity ± SD was ranked by NNs, followed by SVMs, LR, and DT, which were 82.02 ± 8.34%, 80.37 ± 5.15%, 79.65% ± 6.01%, and 78.97% ± 4.07%, respectively.

Conclusion

Both statistics and the ML model could achieve satisfactory predictions for EP. In model learning, the highest ranked model was LR, showing that EP prediction might possess linear or causal data pattern. However, in new testing data, NNs could overcome statistics. This highlights the potency of ML in solving complicated problems with various patterns, while overcoming generalization error of data.

Management of early gestations with low beta-human chorionic gonadotropin conceived by assisted reproductive technologies: performance of M4 predictive model

OBJECTIVES

To assess the safety and performance of the M4 model for classifying as high risk or low risk for ectopic pregnancy (EP) pregnancies conceived by assisted reproductive technologies (ART) that present with low beta-human chorionic gonadotropin (β-hCG) concentration in early gestation.

METHODS

This was a prospective cohort study of 243 pregnancies conceived by ART with low β-hCG levels (5-50 IU/L) at 4 + 0 to 4 + 2 weeks' gestation. After subsequent β-hCG testing at 48 h, pregnancies were classified according to the M4 model into the following categories: (i) high risk, probable EP/persistent pregnancy of unknown location (PPUL), when the risk for EP was ≥ 5%; (ii) low risk, probable intrauterine pregnancy (IUP), when the risk of EP was < 5% and the likelihood of IUP was greater than that of a failed pregnancy of unknown location (FPUL); and (iii) low risk, probable FPUL, when the risk of EP was < 5% and the likelihood of a FPUL was greater than that of an IUP. The predictive performance of the M4 model for EP and its ability to discriminate between high- and low-risk pregnancies was assessed using the final pregnancy outcome at 11 to 13 weeks of gestation as reference, which was classified as EP/PPUL, FPUL or IUP.

RESULTS

The sensitivity and specificity of the M4 model in detecting a high-risk pregnancy (EP/PPUL) were 60.0% (95% CI, 43.6-74.4%) and 79.8% (95% CI, 73.8-84.7%), respectively. The area under the receiver-operating-characteristics curve of the M4 model for discriminating between high-risk and low-risk (FPUL/IUI) pregnancies was 0.72 (95% CI, 0.62-0.81). The model had a positive likelihood ratio of 2.97 (95% CI, 2.03-4.36) and a negative likelihood ratio of 0.50 (95% CI, 0.33-0.76). The kappa index was 0.30 (95% CI, 0.16-0.43), indicating a low degree of agreement between the model classification and the final diagnosis. No serious adverse events related directly to the application of the M4 model were observed, although 14 pregnancies classified ultimately as high risk had been categorized initially as low risk by the M4 model. Of these, seven resolved with expectant management, five with methotrexate (MTX) and two required laparoscopic surgery (one after failure of medical treatment with MTX and one after deviation from the follow-up protocol). There were no cases of EP/PPUL with additional complications or need for blood or other blood product transfusion. Of the 243 ART pregnancies with low β-hCG concentration in early gestation, only 47 (19.3%) had an IUP, half (24/47) of which had an early miscarriage, resulting in only 9.5% (23/243) cases having an ongoing pregnancy.

CONCLUSIONS

Application of the M4 model in pregnancies conceived by ART with low β-hCG concentration in early gestation showed limited capacity in classifying them as being at low or high risk for EP, therefore, its use in pregnancies of this type is not recommended. No serious adverse events or complications related to the use of the model were observed. These pregnancies have a low probability of ending in an IUP as well as a high rate of early miscarriage. © 2021 International Society of Ultrasound in Obstetrics and Gynecology.

Prediction models in gynaecology: Transparent reporting needed for clinical application

The clinical application of prediction models is increasing within the field of gynaecology and obstetrics. This is mostly due to the fact that clinicians and patients prefer individualized counselling and person specific, more objective outcome assessment. To prevent using inadequate models, it is important to construct and perform prediction model studies correctly. Therefore, the TRIPOD statement (the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis) was developed. The aim of this review is to obtain an overview of the existing published prediction models for benign gynaecology and to investigate to what extent these studies meet the TRIPOD criteria. We performed a literature search in the databases PubMed, Embase and Cochrane Library from inception to August 2020. Searching the cross-references of the relevant studies within our search identified additional articles. Publications were included if the aim of the study was to develop a multivariable prediction model within the field of benign gynaecology. Two independent reviewers extracted the data. Analysis of the studies was performed by using a checklist derived from the TRIPOD criteria. Based on our search, 2487 studies were selected, including potential duplications. Eventually, a total of twenty-two studies were selected. 91% of these studies handled their predictors by univariable analysis before developing a multivariable prediction model. Fifteen studies described having missing data, but not all of them (9%) handled these missing data. Four different internal validation methods were used in twenty studies. Fifteen studies (68%) had prediction models with a C-index ≥ 0.7, which indicates a good model. Half of the studies (50%) did not measure the calibration, overall performance was described in two studies (9%). External validation was performed in 9% of the studies. The correct development of a prediction model within benign gynaecology and subsequent transparent reporting of the model development is important to facilitate clinical use. Without transparent reporting, wrong assumptions can be made leading to incorrect application of a specific prediction model. This overview shows that excepting carrying out an external validation, only one article met all the criteria. Therefore, we strongly recommend use of the TRIPOD criteria for developing and validating a prediction model (study). In addition, prior to publication, content experts should critically and statistically review the prediction model. If too many criteria are not met, refusing publication should be considered.

Predictive Analytic Model for Diagnosis of Ectopic Pregnancy

Objective: Ectopic pregnancy (EP) is a serious condition. Delayed diagnosis could lead to life-threatening outcomes. The study aimed to develop a diagnostic predictive model for EP to approach suspected cases with prompt intervention before the rupture occurred.

Methods: A retrospective cross-sectional study enrolled 347 pregnant women presenting first-trimester complications (abdominal pain or vaginal bleeding) with diagnosis suspected of pregnancy of unknown location, who were eligible and underwent chart review. The data including clinical risk factors, signs and symptoms, serum human chorionic gonadotropin (hCG), and ultrasound findings were analyzed. The statistical predictive score was developed by performing logistic regression analysis. The testing data of 30 patients were performed to test the validation of predictive scoring.

Results: From a total of 22 factors, logistic regression method–derived scoring model was based on five potent factors (history of pelvic inflammatory disease, current use of emergency pills, cervical motion tenderness, serum hCG ≥1,000 mIU/ml, and ultrasound finding of adnexal mass) using a cutoff score ≥3. This predictive index score was able to determine ectopic pregnancy with an accuracy of 77.8% [95% confidence interval (CI) = 73.1–82.1], specificity of 91.0% (95% CI = 62.1–72.0), sensitivity of 67.0% (95% CI = 88.0–94.0), and area under the curve of 0.906 (95% CI = 0.875–0.937). In the validation group, no patient with negative result of this score had an EP.

Conclusion: Statistical predictive score was derived with high accuracy and applicable performance for EP diagnosis. This score could be used to support clinical decision making in routine practice for management of EP.

Prediction of Tubal Ectopic Pregnancy Using Offline Analysis of 3-Dimensional Transvaginal Ultrasonographic Data Sets: An Interobserver and Diagnostic Accuracy Study

OBJECTIVES

To assess interobserver reproducibility in detecting tubal ectopic pregnancies by reading data sets from 3-dimensional (3D) transvaginal ultrasonography (TVUS) and comparing it with real-time 2-dimensional (2D) TVUS.

METHODS

Images were initially classified as showing pregnancies of unknown location or tubal ectopic pregnancies on real time 2D TVUS by an experienced sonologist, who acquired 5 3D volumes. Data sets were analyzed offline by 5 observers who had to classify each case as ectopic pregnancy or pregnancy of unknown location. The interobserver reproducibility was evaluated by the Fleiss κ statistic. The performance of each observer in predicting ectopic pregnancies was compared to that of the experienced sonologist. Women were followed until they were reclassified as follows: (1) failed pregnancy of unknown location; (2) intrauterine pregnancy; (3) ectopic pregnancy; or (4) persistent pregnancy of unknown location.

RESULTS

Sixty-one women were included. The agreement between reading offline 3D data sets and the first real-time 2D TVUS was very good (80%-82%; κ = 0.89). The overall interobserver agreement among observers reading offline 3D data sets was moderate (κ = 0.52). The diagnostic performance of experienced observers reading offline 3D data sets had accuracy of 78.3% to 85.0%, sensitivity of 66.7% to 81.3%, specificity of 79.5% to 88.4%, positive predictive value of 57.1% to 72.2%, and negative predictive value of 87.5% to 91.3%, compared to the experienced sonologist's real-time 2D TVUS: accuracy of 94.5%, sensitivity of 94.4%, specificity of 94.5%, positive predictive value of 85.0%, and negative predictive value of 98.1%.

CONCLUSIONS

The diagnostic accuracy of 3D TVUS by reading offline data sets for predicting ectopic pregnancies is dependent on experience. Reading only static 3D data sets without clinical information does not match the diagnostic performance of real time 2D TVUS combined with clinical information obtained during the scan.

Morphological ultrasound types known as 'blob' and 'bagel' signs should be reclassified from suggesting probable to indicating definite tubal ectopic pregnancy

OBJECTIVE

In a recent consensus statement on early pregnancy nomenclature by Barnhart, a definite ectopic pregnancy (EP) was defined morphologically on transvaginal sonography (TVS) as an extrauterine gestational sac with yolk sac and/or embryo, with or without cardiac activity, whilst a probable EP was defined as an inhomogeneous adnexal mass ('blob' sign) or extrauterine sac-like structure ('bagel' sign). This study aims to determine whether these ultrasound markers used to define probable EP can be used to predict a definite tubal EP.

METHODS

This was a retrospective cohort study of women presenting to the Early Pregnancy Unit (EPU) at Nepean Hospital, Sydney, Australia between November 2006 and June 2016. Women classified with a probable EP or a pregnancy of unknown location (PUL), i.e. with no signs of extra- or intrauterine pregnancy (IUP), at their first TVS were included, whilst those with a definite tubal EP, IUP or non-tubal EP were excluded from the final analysis. The gold standard for tubal EP was histological confirmation of chorionic villi in Fallopian tube removed at laparoscopy. The performance of blob or bagel sign on TVS in the prediction of definite tubal EP was evaluated in terms of sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV). This was compared with the performance of extrauterine gestational sac with yolk sac and/or embryo on TVS to predict definite tubal EP.

RESULTS

During the study period, 7490 consecutive women attended the EPU, of whom 849 were analyzed. At primary TVS, 240/849 were diagnosed with probable EP, of which 174 (72.5%) were classified as blob sign and 66 (27.5%) as bagel sign. The remaining 609/849 were diagnosed with PUL, of which 47 had a final diagnosis of EP (including 24 blob sign, 19 bagel sign and four gestational sac with embryo/yolk sac). 101 of all 198 (51%) blob sign cases and 50 of all 85 (59%) bagel sign cases underwent laparoscopy and salpingectomy; histology proved a tubal EP in 98 (97%) of these blob-sign cases and 48 (96.0%) of the bagel-sign cases. The sensitivity for the blob and bagel signs in the prediction of definite tubal EP was 89.8% and 83.3%, respectively, the specificity was 99.5% and 99.6%, PPV was 96.7% and 95.2% and NPV was 98.3% and 98.6%. This was comparable to the sensitivity of extrauterine gestational sac with yolk sac and/or embryo on TVS in the prediction of definite tubal EP (sensitivity, 84.0%; specificity, 99.9%; PPV, 97.7%; NPV, 99.3% (P = 0.5)).

CONCLUSIONS

Blob and bagel signs seem to be the most common presentations of a tubal EP on TVS. Although they cannot be considered as a definitive sign of EP, their PPV is very high (> 95%); such women should therefore be considered at very high risk for having a tubal EP and should be treated as such. Copyright © 2017 ISUOG. Published by John Wiley & Sons Ltd.

Association between three-dimensional transvaginal sonographic markers and outcome of pregnancy of unknown location: a pilot study

OBJECTIVE

To assess the accuracy of three-dimensional (3D) transvaginal sonographic (TVS) parameters in predicting the evolution of a pregnancy of unknown location (PUL).

METHODS

This was a prospective observational study performed at the early pregnancy unit of a university hospital from September 2008 to June 2012. Women with a positive pregnancy test without any signs of intra- or extrauterine pregnancy at their first TVS examination were considered eligible and a 3D dataset containing the entire uterus was acquired. An experienced observer analyzed all 3D datasets for assessment of the following parameters: endometrial thickness, volume, mean gray-scale index and asymmetry. Women were followed until they were classified as having: (i) non-visualized pregnancy loss (NVPL); (ii) intrauterine pregnancy (IUP); or (iii) ectopic pregnancy or persistent PUL. We compared the values of the TVS parameters across the three groups. We also assessed the area under the receiver-operating characteristics curve of the 3D-TVS parameters in comparison to that for serum β-human chorionic gonadotropin (β-hCG) ratio (48 h/baseline) to predict PUL outcome. We then evaluated whether combining the 3D-TVS parameters with serum β-hCG ratio improved the predictive accuracy for PUL outcome by performing a logistic regression analysis.

RESULTS

During the study period 4939 consecutive pregnant women presented at the unit for their initial TVS examination and 325 (7%) were classified as having a PUL, of whom 161 women were enrolled and had a 3D scan of the uterus. However, 19 were excluded because of incomplete follow-up. Data from 142 women with PUL were therefore included in the analysis and the outcomes of these women were: NVPL in 98 (69%), IUP in 27 (19%) and ectopic pregnancy + persistent PUL in 14 + 3 = 17 (12%). Endometrial thickness, endometrial volume and the proportion of women with asymmetric endometrial shape differed significantly between the outcome groups. Endometrial thickness and volume could be used as reasonable predictors of both NVPL and IUP, whereas asymmetric endometrial shape and mean gray-scale index could be used as reasonable predictors of IUP only. The best single parameter to predict PUL outcomes was the β-hCG ratio. Regression analysis demonstrated that endometrial volume and endometrial shape asymmetry added significantly to the β-hCG ratio in predicting IUP but not NVPL.

CONCLUSIONS

3D-TVS markers have a low diagnostic accuracy in predicting PUL outcome. The addition of endometrial volume and shape asymmetry improves the accuracy of the β-hCG ratio in predicting IUP. Copyright © 2016 ISUOG. Published by John Wiley & Sons Ltd.

Managing pregnancy of unknown location based on initial serum progesterone and serial serum hCG levels: development and validation of a two-step triage protocol

OBJECTIVES

A uniform rationalized management protocol for pregnancies of unknown location (PUL) is lacking. We developed a two-step triage protocol to select PUL at high risk of ectopic pregnancy (EP), based on serum progesterone level at presentation (step 1) and the serum human chorionic gonadotropin (hCG) ratio, defined as the ratio of hCG at 48 h to hCG at presentation (step 2).

METHODS

This was a cohort study of 2753 PUL (301 EP), involving a secondary analysis of prospectively and consecutively collected PUL data from two London-based university teaching hospitals. Using a chronological split we used 1449 PUL for development and 1304 for validation. We aimed to assign PUL as low risk with high confidence (high negative predictive value (NPV)) while classifying most EP as high risk (high sensitivity). The first triage step assigned PUL as low risk using a threshold of serum progesterone at presentation. The remaining PUL were triaged using a novel logistic regression risk model based on hCG ratio and initial serum progesterone (second step), defining low risk as an estimated EP risk of < 5%.

RESULTS

On validation, initial serum progesterone ≤ 2 nmol/L (step 1) classified 16.1% PUL as low risk. Second-step classification with the risk model selected an additional 46.0% of all PUL as low risk. Overall, the two-step protocol classified 62.1% of PUL as low risk, with an NPV of 98.6% and a sensitivity of 92.0%. When the risk model was used in isolation (i.e. without the first step), 60.5% of PUL were classified as low risk with 99.1% NPV and 94.9% sensitivity.

CONCLUSION

PUL can be classified efficiently into being either high or low risk for complications using a two-step protocol involving initial progesterone and hCG levels and the hCG ratio. Copyright © 2016 ISUOG. Published by John Wiley & Sons Ltd.

The clinical performance of the M4 decision support model to triage women with a pregnancy of unknown location as at low or high risk of complications

STUDY QUESTION

What are the adverse outcomes associated with using the M4 model in everyday clinical practice for women with pregnancy of unknown location (PUL)?

SUMMARY ANSWER

There were 17/835 (2.0%) adverse events and no serious adverse events associated with the performance of the M4 model in clinical practice.

WHAT IS KNOWN ALREADY

The M4 model has previously been shown to stratify women classified as a PUL as at low or high risk of complications with a good level of test performance. The triage performance of the M4 model is better than single measurements of serum progesterone or the hCG ratio (serum hCG at 48 h/hCG at presentation).

STUDY DESIGN, SIZE, DURATION

A prospective multi-centre cohort study of 1022 women with a PUL carried out between August 2012 and December 2013 across 2 university teaching hospitals and 1 district general hospital.

PARTICIPANTS/MATERIALS, SETTING, METHODS

All women presenting with a PUL to the early pregnancy units of the three hospitals were recruited. The final outcome for PUL was either a failed PUL (FPUL), intrauterine pregnancy (IUP) or ectopic pregnancy (EP) (including persistent PUL (PPUL)), with EP and PPUL considered high-risk PUL. Their hCG results at 0 and 48 h were entered into the M4 model algorithm. If the risk of EP was ≥5%, the PUL was predicted to be high-risk and the participant was asked to re-attend 48 h later for a repeat hCG and transvaginal ultrasound scan by a senior clinician. If the PUL was classified as 'low risk, likely failed PUL', the participant was asked to perform a urinary pregnancy test 2 weeks later. If the PUL was classified as 'low risk, likely intrauterine', the participant was scheduled for a repeat scan in 1 week. Deviations from the management protocol were recorded as either an 'unscheduled visit (participant reason)', 'unscheduled visit (clinician reason)' or 'differences in timing (blood test/ultrasound)'. Adverse events were assessed using definitions outlined in the UK Good Clinical Practice Guidelines' document.

MAIN RESULTS AND THE ROLE OF CHANCE

A total of 835 (82%) women classified as a PUL were managed according to the M4 model (9 met the exclusion criteria, 69 were lost to follow-up, 109 had no hCG result at 48 h). Of these, 443 (53%) had a final outcome of FPUL, 298 (36%) an IUP and 94 (11%) an EP. The M4 model predicted 70% (585/835) PUL as low risk, of which 568 (97%) were confirmed as FPUL or IUP. Of the 17 EP and PPUL misclassified as low risk, 5 had expectant management, 7 medical management with methotrexate and 5 surgical intervention.Nineteen PUL had an unscheduled visit (participant reason), 38 PUL had an unscheduled visit (clinician reason) and 68 PUL had deviations from protocol due to a difference in timing (blood test/ultrasound).Adverse events were reported in 26 PUL and 1 participant had a serious adverse event. A total of 17/26 (65%) adverse events were misclassifications of a high risk PUL as low risk by the M4 model, while 5/26 (19%) adverse events were related to incorrect clinical decisions. Four of the 26 adverse events (15%) were secondary to unscheduled admissions for pain/bleeding. The serious adverse event was due to an incorrect clinical decision.

LIMITATIONS, REASONS FOR CAUTION

A limitation of the study was that 69/1022 (7%) of PUL were lost to follow-up. A 48 h hCG level was missing for 109/1022 (11%) participants.

WIDER IMPLICATIONS OF THE FINDINGS

The low number of adverse events (2.0%) suggests that expectant management of PUL using the M4 prediction model is safe. The model is an effective way of triaging women with a PUL as being at high- and low-risk of complications and rationalizing follow-up. The multi-centre design of the study is more likely to make the performance of the M4 model generalizable in other populations.

STUDY FUNDING/COMPETING INTERESTS

None.

TRIAL REGISTRATION NUMBER

Not applicable.

Molecular diagnostics and therapeutics for ectopic pregnancy

Ectopic pregnancies are a serious gynaecological emergency that can be fatal. As such, prompt diagnosis and safe timely treatment is essential. Here, we review the literature on the development of molecularly targeted diagnostics and therapeutics for ectopic pregnancy. A blood-based biomarker that accurately identifies an ectopic pregnancy could be used to offer early diagnostic certainty in cases where ultrasound cannot determine the location of the embryo ('a pregnancy of unknown location'). Molecules examined so far can be broadly grouped into biological themes of relevance to reproduction: (i) Fallopian tube (dys)function, (ii) embryo/trophoblast growth, (iii) corpus luteum function, (iv) inflammation, (v) uterine function and (vi) angiogenesis. While a sensitive and specific biomarker for ectopic pregnancy has yet to be identified, it is possible that improvements in platform technologies or a multi-modal biomarker approach may yield an accurate diagnostic biomarker test. Furthermore, with the advent of better imaging technology, the need for a blood-based biomarker test may be superseded by improvements in ultrasound or magnetic resonance imaging technology. There have been some recent preclinical studies describing molecularly targeted therapeutic approaches for ectopic pregnancy. Notably, bench-to-bedside studies have examined the use of combination gefitinib (orally available epidermal growth factor receptor inhibitor) and methotrexate. Preclinical studies suggest that combination gefitinib and methotrexate is highly effective in inducing placental cell death, and is significantly more effective than methotrexate alone. In early human trials, encouraging preliminary efficacy data have shown that combination gefitinib and methotrexate can rapidly resolve tubal ectopic pregnancies, and large extra-tubal ectopic pregnancies. If a large clinical randomized controlled trial confirms these findings, combination gefitinib and methotrexate could become a new medical treatment option for ectopic pregnancy.

MicroRNA and implantation

We provide a review of microRNA (miRNA) related to human implantation which shows the potential diagnostic role of miRNAs in impaired endometrial receptivity, altered embryo development, implantation failure after assisted reproduction technology, and in ectopic pregnancy and pregnancies of unknown location. MicroRNAs may be emerging diagnostic markers and potential therapeutic tools for understanding implantation disorders. However, further research is needed before miRNAs can be used in clinical practice for identifying and treating implantation failure.

Ectopic pregnancy prediction in women with a pregnancy of unknown location: data beyond 48 h are necessary

STUDY QUESTION

Are there improvements in the accuracy of prediction of ectopic pregnancy (EP) in women with early symptomatic pregnancy using human chorionic gonadotrophin (hCG) curves when clinicians consider visits beyond the first 48 h after initial presentation?

SUMMARY ANSWER

Two hCG values, measured 48 h (2 days) apart, are often not sufficient to accurately predict the outcome of a woman with a pregnancy of unknown location (PUL), but adding a third visit on Day 4 or 7 significantly improved the prediction for 1 in 15 women.

WHAT IS KNOWN ALREADY

The use of serial hCG values is commonly used to aid in the prediction of the final diagnosis in women with a PUL. Initial outcome predictions based on two hCG values may often be incorrect.

STUDY DESIGN, SIZE, DURATION

This retrospective multicenter cohort study included 646 women with a PUL, recruited over 2 years. Of these women, 146 were ultimately diagnosed with EP.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Women presenting to the emergency room with first trimester pain or bleeding, with a PUL, at least 2 hCG values and a definitive final diagnosis from the University of Pennsylvania, University of Miami and University of Southern California, were recruited from 2007 to 2009.

MAIN RESULTS AND THE ROLE OF CHANCE

Using currently recommended prediction rules, adding a third hCG evaluation on Day 4 after initial presentation significantly improved the accuracy of initial prediction from the first two values (48 h apart, or Day 2) by 9.3% (P = 0.015). Adding a third value on Day 7 improved prediction significantly by 6.7% (P = 0.031), compared with prediction based on first two values. The improvement in prediction by assessing four hCG values (Days 0, 2, 4 and 7) compared with three values (Days 0, 2 and 4) was 1.3% and not statistically significant.

LIMITATIONS, REASONS FOR CAUTION

Missing data imputation likely biased results toward the null; predicted outcomes may not match those made by clinicians; and the study does not predict intrauterine pregnancy and spontaneous miscarriage separately.

WIDER IMPLICATIONS OF THE FINDINGS

This study provides useful information for the prediction of outcomes for women with a symptomatic first trimester pregnancy of unknown location, but may not be generalizable to all pregnant women.

STUDY FUNDING/COMPETING INTEREST(S)

Supported by NIH grant numbers R01-HD036455 to Dr Barnhart and Dr Sammel, K24HD060687 to Dr Barnhart, and 5T32MH065218 to Ms. Zee. The authors have no conflicts of interest to declare.

Shotgun proteomics identifies serum fibronectin as a candidate diagnostic biomarker for inclusion in future multiplex tests for ectopic pregnancy

Ectopic pregnancy (EP) is difficult to diagnose early and accurately. Women often present at emergency departments in early pregnancy with a ‘pregnancy of unknown location’ (PUL), and diagnosis and exclusion of EP is challenging due to a lack of reliable biomarkers. The objective of this study was to identify novel diagnostic biomarkers for EP. Shotgun proteomics, incorporating combinatorial-ligand library pre-fractionation, was used to interrogate pooled sera (n = 40) from women undergoing surgery for EP, termination of viable intrauterine pregnancy and management of non-viable intrauterine pregnancy. Western blot was used to validate results in individual sera. ELISAs were developed to interrogate sera from women with PUL (n = 120). Sera were collected at time of first symptomatic presentation and categorized according to pregnancy outcome. The main outcome measures were differences between groups and area under the receiver operating curve (ROC). Proteomics identified six biomarker candidates. Western blot detected significant differences in levels of two of these candidates. ELISA of sera from second cohort revealed that these differences were only significant for one of these candidates, fibronectin. ROC analysis of ability of fibronectin to discriminate EP from other pregnancy outcomes suggested that fibronectin has diagnostic potential (ROC 0.6439; 95% CI 0.5090 to 0.7788; P>0.05), becoming significant when ‘ambiguous’ medically managed PUL excluded from analysis (ROC 0.6538; 95% CI 0.5158 to 0.7918; P<0.05). Fibronectin may make a useful adjunct to future multiplex EP diagnostic tests.

Biomarkers for ectopic pregnancy and pregnancy of unknown location

Early pregnancy failure is the most common complication of pregnancy, and 1% to 2% of all pregnancies will be ectopic. As one of the leading causes of maternal morbidity and mortality, diagnosing ectopic pregnancy and determining the fate of a pregnancy of unknown location are of great clinical concern. Several serum and plasma biomarkers for ectopic pregnancy have been investigated independently and in combination. The following is a review of the state of biomarker discovery and development for ectopic pregnancy and pregnancy of unknown location.

Evaluation of ADAM-12 as a diagnostic biomarker of ectopic pregnancy in women with a pregnancy of unknown location

Background

Ectopic pregnancy (EP) remains the most life-threatening acute condition in modern gynaecology. It remains difficult to diagnose early and accurately. Women often present at emergency departments in early pregnancy with a ‘pregnancy of unknown location’ (PUL) and diagnosis/exclusion of EP is challenging due to a lack of reliable biomarkers. Recent studies suggest that serum levels of a disintegrin and metalloprotease protein-12 (ADAM-12) can be used differentiate EP from viable intrauterine pregnancy (VIUP). Here we describe a prospective study evaluating the performance of ADAM-12 in differentiating EP from the full spectrum of alternative PUL outcomes in an independent patient cohort.

Methodology/Principal Findings

Sera were collected from 120 patients at their first clinical presentation with a PUL and assayed for ADAM-12 by ELISA. Patients were categorized according to final pregnancy outcomes. Serum ADAM-12 concentrations were increased in women with histologically-confirmed EP (median 442 pg/mL; 25%–75% percentile 232–783 pg/mL) compared to women with VIUP (256 pg/mL; 168–442 pg/mL) or miscarriage (192 pg/mL; 133–476 pg/mL). Serum ADAM-12 did not differentiate histologically-confirmed EP from spontaneously resolving PUL (srPUL) (416 pg/mL; 154–608 pg/mL). The diagnostic potential of ADAM-12 was only significant when ‘ambiguous’ PUL outcomes were excluded from the analysis (AROC = 0.6633; P = 0.03901).

Conclusions/Significance

When measured in isolation, ADAM-12 levels had limited value as a diagnostic biomarker for EP in our patient cohort. The development of a reliable serum biomarker-based test for EP remains an ongoing challenge.

The diagnostic value of a multivariate logistic regression analysis model with transvaginal power Doppler ultrasonography for the prediction of ectopic pregnancy

OBJECTIVE

A multivariate logistic regression analysis model for predicting ectopic pregnancy in women with pregnancy of unknown location was designed and evaluated clinically.

METHODS

Endometrial thickness, symmetry, resonance, pattern of echogenicity, helicine artery blood flow and blood flow resistance index (RI) in 129 patients with suspected early ectopic pregnancy were assessed by transvaginal power Doppler ultrasonography. Variables significant in univariate logistic regression analysis were included in a multivariate predictive logistic regression analysis model.

RESULTS

The final predictive model included three factors: endometrial thickness≤9 mm; a multilayered endometrial echogenicity pattern with prominent outer and midline hyperechogenic lines and an inner hypoechogenic region; and visible endometrial arterial blood flow. The area under the receiver operating characteristic curve of the model was 0.980. When RI was >0.65 and the predictive probability>0.50, diagnostic accuracy was high. The model correctly diagnosed 52/55 (94.5%) clinically confirmed ectopic pregnancy cases.

CONCLUSION

This multivariate predictive logistic regression analysis model has clinical value for the differential diagnosis of early ectopic pregnancy when the pregnancy location is unknown.

Performance of human chorionic gonadotropin curves in women at risk for ectopic pregnancy: exceptions to the rules

OBJECTIVE

To investigate the accuracy of serial hCG to predict outcome of a pregnancy of unknown location in an ethnically and geographically diverse setting.

DESIGN

Multisite cohort study.

SETTING

University hospital.

PATIENT(S)

Women with a pregnancy of unknown location.

INTERVENTION(S)

None.

MAIN OUTCOME MEASURE(S)

Patients were followed until diagnosed with ectopic pregnancy (EP), intrauterine pregnancy (IUP), or miscarriage. To predict outcome, observed hCG level was compared with recommended thresholds to assess deviation from defined normal curves. Predicted outcome was compared with standard of care. Sensitivity, specificity, predictive value, and accuracy were calculated, stratified by diagnosis.

RESULT(S)

The final diagnosis of 1,005 patients included 179 EPs, 259 IUPs, and 567 miscarriages. The optimal balance in sensitivity and specificity used the minimal expected 2-day increase in hCG level of 35%, and the minimal 2-day decrease in hCG level of 36%-47% (depending on the level) achieving 83.2% sensitivity, 70.8% specificity to predict EP. However, 16.8% of EPs and 7.7% of IUPs would be misclassified solely using serial hCG levels. Consideration of a third hCG and early ultrasound decreased IUP misclassification to 2.7%.

CONCLUSION(S)

Solely using serial hCG values can result in misclassification. Clinical judgment should trump prediction rules and continued surveillance with a third hCG may be prudent, especially when initial values are low or when values are near suggested thresholds.

Improved modeling of clinical data with kernel methods

OBJECTIVE

Despite the rise of high-throughput technologies, clinical data such as age, gender and medical history guide clinical management for most diseases and examinations. To improve clinical management, available patient information should be fully exploited. This requires appropriate modeling of relevant parameters.

METHODS

When kernel methods are used, traditional kernel functions such as the linear kernel are often applied to the set of clinical parameters. These kernel functions, however, have their disadvantages due to the specific characteristics of clinical data, being a mix of variable types with each variable its own range. We propose a new kernel function specifically adapted to the characteristics of clinical data.

RESULTS

The clinical kernel function provides a better representation of patients' similarity by equalizing the influence of all variables and taking into account the range r of the variables. Moreover, it is robust with respect to changes in r. Incorporated in a least squares support vector machine, the new kernel function results in significantly improved diagnosis, prognosis and prediction of therapy response. This is illustrated on four clinical data sets within gynecology, with an average increase in test area under the ROC curve (AUC) of 0.023, 0.021, 0.122 and 0.019, respectively. Moreover, when combining clinical parameters and expression data in three case studies on breast cancer, results improved overall with use of the new kernel function and when considering both data types in a weighted fashion, with a larger weight assigned to the clinical parameters. The increase in AUC with respect to a standard kernel function and/or unweighted data combination was maximum 0.127, 0.042 and 0.118 for the three case studies.

CONCLUSION

For clinical data consisting of variables of different types, the proposed kernel function--which takes into account the type and range of each variable--has shown to be a better alternative for linear and non-linear classification problems.

Methotrexate treatment of ectopic pregnancy: experience at nizwa hospital with literature review

OBJECTIVES

Ectopic pregnancy occurs when the embryo fails to implant within the uterine cavity. Methotrexate, a folic acid antagonist has been widely used to treat ectopic pregnancy. The aim of this study is to determine the efficacy of methotrexate treatment for ectopic pregnancies in our settings which will in-turn help us reduce maternal mortality and morbidity.

METHODS

This was a retrospective review of 60 cases of ectopic pregnancy treated with single dose methotrexate in Nizwa hospital, Oman. Demographic data, clinical presentation, treatment progress, outcome, side effects and future fertility follow up were analysed using a computer database. Selection criteria included hemodynamically stable women with ß-hcg level of ≤5000 mIU/ml, adnexal mass ≤4 cm, absent cardiac activity and hemoperitoneum less than 100 ml. Human chorionic gonadotrophin (hcg) assay and clinical presentation were used to determine cases which warranted surgical intervention.

RESULTS

The success rate of single dose methotrexate therapy in our study was 65% (n=39) and 35% (n=21) required surgical intervention compared to reported success rate of 67-100% published in various studies. Careful selection of cases and confining strictly to the criteria improved the success rate from 40% to 60% in five and half years of the study. The mean average time of resolution of ectopic pregnancy was 32 days for a single dose and 58 days for repeat second or third doses. Eight cases of pregnancy of unknown location were treated successfully with Methotrexate. No major side effects were noted. A successful fertility outcome of 30% in the first year followed by 13.3% in the subsequent second year with a secondary infertility rate of 11.6% was observed in this study.

CONCLUSION

Methotrexate treatment of ectopic pregnancies is safe and effective with no major side effects. Intramuscular methotrexate has the advantage of tubal conservation and saves patients from requiring surgery. It is easier to administer than intraoperative route, which is by laparoscopy and needs expertise. Our study showed single dose methotrexate to be an effective treatment option for selected patients with unruptured tubal ectopic pregnancy.

The term "pregnancy of unknown location" is here to stay

The term "pregnancy of unknown location" is an ultrasound classification and not a final diagnosis. The use of this terminology is here to stay and should continue as long as there is an appreciation for what it really means. It is the responsibility of the clinician, who follows up these women with a PUL, to ensure that a final diagnosis is achieved while preserving the well-being of these women.

Predictive power of activin A levels in the prognosis of first trimester in vitro fertilization pregnancies

BACKGROUND

The present study aims to determine the predictive power of activin A levels in the prognosis of first trimester pregnancies conceived by in vitro fertilization (IVF).

METHODS

The study cohort included 23 biochemical, and 23 normal ongoing pregnancies conceived via IVF. Serum β-human chorionic gonadotropin (β-hCG), progesterone, and activin A levels were assessed 14 days after embryo transfer.

RESULTS

Serum activin A levels were significantly lower in biochemical pregnancies compared to normal ongoing pregnancies (0.57 vs. 0.81 ng/mL, p<0.001). The ability of activin A to predict normal ongoing pregnancies at a cutoff level of 0.695 ng/mL gave a sensitivity of 91.3%, a specificity of 100%, a positive predictive value of 100%, and a negative predictive value of 92%. Significant correlation was found between β-hCG and progesterone, β-hCG and activin A, and progesterone and activin A. Activin A, β-hCG, and progesterone were all found to be efficacious in prediction of early IVF pregnancies.

CONCLUSIONS

The present study indicates that single measurement of activin A can be suggested as a marker for the progress and outcome of early pregnancies conceived via IVF. However, further large-scale studies are required to determine the efficacy and reliability of activin A in prediction of early pregnancies achieved by assisted reproductive techniques (ART).

Pregnancy of unknown location: a consensus statement of nomenclature, definitions, and outcome

OBJECTIVE

To improve the interpretation of future studies in women who are initially diagnosed with a pregnancy of unknown location (PUL), we propose a consensus statement with definitions of population, target disease, and final outcome.

DESIGN

A review of literature and a series of collaborative international meetings were used to develop a consensus for definitions and final outcomes of women initially diagnosed with a PUL.

RESULT(S)

Global differences were noted in populations studied and in the definitions of outcomes. We propose to define initial ultrasound classification of findings into five categories: definite ectopic pregnancy (EP), probable EP, PUL, probable intrauterine pregnancy (IUP), and definite IUP. Patients with a PUL should be followed and final outcomes should be categorized as visualized EP, visualized IUP, spontaneously resolved PUL, and persisting PUL. Those with the transient condition of a persisting PUL should ultimately be classified as nonvisualized EP, treated persistent PUL, resolved persistent PUL, or histologic IUP. These specific categories can be used to characterize the natural history or location (intrauterine vs. extrauterine) of any early gestation where the initial location is unknown.

CONCLUSION(S)

Careful definition of populations and classification of outcomes should optimize objective interpretation of research, allow objective assessment of future reproductive prognosis, and hopefully lead to improved clinical care of women initially identified to have a PUL.

Does a prediction model for pregnancy of unknown location developed in the UK validate on a US population?

BACKGROUND

A logistic regression model (M4) was developed in the UK to predict the outcome for women with a pregnancy of unknown location (PUL) based on the initial two human chorionic gonadotrophin (hCG) values, 48 h apart. The purpose of this paper was to assess the utility of this model to predict the outcome for a woman (PUL) in a US population.

METHODS

Diagnostic variables included log-transformed serum hCG average of two measurements, and linear and quadratic hCG ratios. Outcomes modeled were failing PUL, intrauterine pregnancy (IUP) and ectopic pregnancy (EP). This model was applied to a US cohort of 604 women presenting with symptomatic first-trimester pregnancies, who were followed until a definitive diagnosis was made. The model was applied before and after correcting for differences in terminology and diagnostic criteria.

RESULTS

When retrospectively applied to the adjusted US population, the M4 model demonstrated lower areas under the curve compared with the UK population, 0.898 versus 0.988 for failing PUL/spontaneous miscarriage, 0.915 versus 0.981 for IUP and 0.831 versus 0.904 for EP. Whereas the model had 80% sensitivity for EP using UK data, this decreased to 49% for the US data, with similar specificities. Performance only improved slightly (55% sensitivity) when the US population was adjusted to better match the UK diagnostic criteria.

CONCLUSIONS

A logistic regression model based on two hCG values performed with modest decreases in predictive ability in a US cohort for women at risk for EP compared with the original UK population. However, the sensitivity for EP was too low for the model to be used in clinical practice in its present form. Our data illustrate the difficulties of applying algorithms from one center to another, where the definitions of pathology may differ.

L2-norm multiple kernel learning and its application to biomedical data fusion

BACKGROUND

This paper introduces the notion of optimizing different norms in the dual problem of support vector machines with multiple kernels. The selection of norms yields different extensions of multiple kernel learning (MKL) such as L(infinity), L1, and L2 MKL. In particular, L2 MKL is a novel method that leads to non-sparse optimal kernel coefficients, which is different from the sparse kernel coefficients optimized by the existing L(infinity) MKL method. In real biomedical applications, L2 MKL may have more advantages over sparse integration method for thoroughly combining complementary information in heterogeneous data sources.

RESULTS

We provide a theoretical analysis of the relationship between the L2 optimization of kernels in the dual problem with the L2 coefficient regularization in the primal problem. Understanding the dual L2 problem grants a unified view on MKL and enables us to extend the L2 method to a wide range of machine learning problems. We implement L2 MKL for ranking and classification problems and compare its performance with the sparse L(infinity) and the averaging L1 MKL methods. The experiments are carried out on six real biomedical data sets and two large scale UCI data sets. L2 MKL yields better performance on most of the benchmark data sets. In particular, we propose a novel L2 MKL least squares support vector machine (LSSVM) algorithm, which is shown to be an efficient and promising classifier for large scale data sets processing.

CONCLUSIONS

This paper extends the statistical framework of genomic data fusion based on MKL. Allowing non-sparse weights on the data sources is an attractive option in settings where we believe most data sources to be relevant to the problem at hand and want to avoid a "winner-takes-all" effect seen in L(infinity) MKL, which can be detrimental to the performance in prospective studies. The notion of optimizing L2 kernels can be straightforwardly extended to ranking, classification, regression, and clustering algorithms. To tackle the computational burden of MKL, this paper proposes several novel LSSVM based MKL algorithms. Systematic comparison on real data sets shows that LSSVM MKL has comparable performance as the conventional SVM MKL algorithms. Moreover, large scale numerical experiments indicate that when cast as semi-infinite programming, LSSVM MKL can be solved more efficiently than SVM MKL.

AVAILABILITY

The MATLAB code of algorithms implemented in this paper is downloadable from http://homes.esat.kuleuven.be/~sistawww/bioi/syu/l2lssvm.html.

Development of a kernel function for clinical data

For most diseases and examinations, clinical data such as age, gender and medical history guides clinical management, despite the rise of high-throughput technologies. To fully exploit such clinical information, appropriate modeling of relevant parameters is required. As the widely used linear kernel function has several disadvantages when applied to clinical data, we propose a new kernel function specifically developed for this data. This "clinical kernel function" more accurately represents similarities between patients. Evidently, three data sets were studied and significantly better performances were obtained with a Least Squares Support Vector Machine when based on the clinical kernel function compared to the linear kernel function.

Prediction of outcome in women with symptomatic first-trimester pregnancy: focus on intrauterine rather than ectopic gestation

OBJECTIVE

Symptoms of vaginal bleeding and abdominal pain are common in cases of ectopic pregnancy (EP), spontaneous abortions (SAB), and complications of an intrauterine pregnancy (IUP). It is important to determine if efforts should focus on differentiating EP from an IUP (IUP + SAB) or a viable IUP from a nonviable gestation (EP + SAB) in women at risk for EP.

METHODS

This is a retrospective cohort study of women who presented with bleeding or pain or both during the first trimester of pregnancy. The cohort was divided into subjects diagnosed with IUP vs. (EP + SAB). The same cohort was then divided into subjects diagnosed with EP vs. (IUP + SAB). Logistic regression models based on risk factors for both outcomes (EP vs. [IUP + SAB] and IUP vs. [EP + SAB]) were obtained. ROC curves as well as Hosmer-Lemeshow goodness of fit and Akaike's information criterion (AIC) were used.

RESULTS

Overall, 18.1% (n = 367) of the women were diagnosed with EP, 58.8% (n = 1192) were diagnosed with an SAB, and 23.1% (n = 467) had an ongoing IUP. The area under the ROC curve for the model IUP vs. (EP + SAB) was statistically greater than the model EP vs. (IUP + SAB), p < 0.001. AIC and Hosmer-Lemeshow goodness of fit confirmed the better accuracy of the model comparing IUP vs. (EP + SAB).

CONCLUSIONS

Information collected at initial presentation from women at risk for EP to be used for building prediction rules should focus on differentiating a viable from a nonviable pregnancy rather than attempting to distinguish an extrauterine from an intrauterine pregnancy. However, this distinction should not affect current clinical care.

Does tubal ectopic pregnancy with hemoperitoneum always require surgery?

OBJECTIVE

Hemoperitoneum is accepted as an indication for surgery in women with tubal ectopic pregnancy. The aim of this pilot study was to evaluate the feasibility of managing such women non-surgically.

METHODS

This was a prospective observational study. Women with tubal ectopic pregnancy and hemoperitoneum detected on transvaginal sonography (TVS) were managed as inpatients either expectantly or with methotrexate (MTX). Inclusion criteria for conservative management were: compliance, clinical stability, absence of acute abdomen, stable hemoglobin level on two measurements (0 and 12-24 h apart), serum human chorionic gonadotropin (hCG) < 5000 IU/L, absence of fetal cardiac activity on TVS and absence of significant hemoperitoneum, defined as blood above the level of the uterine fundus and/or in Morison's pouch (hepatorenal space). Subsequent management was based upon the hCG ratio at 48 h. All the women were managed as inpatients until the abdominal pain settled and the serum hCG levels were falling.

RESULTS

Forty-one women with tubal ectopic pregnancy presented between November 2006 and March 2008. Eight women (20%) fulfilled the entry criteria. The median gestational age at diagnosis was 49 (interquartile range, 38-52.5) days. All women presented with lower abdominal pain/right iliac fossa or left iliac fossa pain. Hemoglobin levels ranged from 11.2 to 14.2 g/dL at presentation and from 12.0 to 14.8 g/dL after 12-24 h. 6/8 (75%) women were managed expectantly and 2/8 (25%) received MTX. All women had resolution of their ectopic pregnancy within 3 weeks with no complications.

CONCLUSIONS

This pilot study suggests that the finding of hemoperitoneum on ultrasound examination may not be an absolute contraindication to conservative management of tubal ectopic pregnancy.