[Show abstract][Hide abstract] ABSTRACT: Continuous subcutaneous insulin infusion (CSII) is increasingly being used worldwide, mostly thanks to technical improvements. This study examined the current status of CSII in Italy. Physicians in charge of 272 diabetes centers caring for patients using CSII were sent a questionnaire investigating clinical features, pump technology and management of these patients; a large proportion (217 centers, 79.8%) joined the study. By end-April 2013, data had been collected on 10152 patients treated with CSII; 98.2% had type 1 diabetes, 82.4% were adults, 57% female. Only just over half the centers (59%) managed more than 20 CSII patients each. The distribution of patients varied widely both among and within different regions. The main indication for CSII was the de- sire to improve glycemic control. Dropouts (8.65%) were mainly due to difficulties with pump wearability or non-optimal glycemic control. Among CSII patients 61% used a traditional pump, 39% a sensor augmented pump. Only 68% used the CSII advanced functions and glucose sensors were used twelve days per month on average. Round-the-clock assistance was guaranteed in 81% of centers; a full diabetes team followed patients in only 40% of adult-care centers and 50% of pediatric units. CSII is increasingly used in Italy, by adults and pediatric patients. However, further work is needed to unify treatment strategies throughout the country and to encourage optimal pump use and applications.
Giornale Italiano di Diabetologia e Metabolismo 06/2015; 35(2):121-129.
[Show abstract][Hide abstract] ABSTRACT: Objective Aim of this study was to study the efficacy and safety of long-acting insulin analog Insulin Lispro Protamine (ILPS) in diabetic pregnant women.
In a multicenter observational retrospective study, we evaluated pregnancy outcome in 119 women affected by type 1 diabetes and 814 with gestational diabetes (GDM) treated during pregnancy with ILPS, compared with a control group treated with NPH insulin. Results Among type 1 diabetic patients, fasting blood glucose at the end of pregnancy was significantly lower in ILPS-treated than in NPH-treated patients. HbA1c levels across pregnancy did not differ between groups. Caesarean section and preterm delivery rates were significantly lower in the ILPS-women. Fetal outcomes were similar in the ILPS and NPH groups. Among GDM women, fasting blood glucose at the end of pregnancy was significantly lower in ILPS-treated than in NPH-treated patients. Duration of gestation was significantly longer, Caesarian section and preterm delivery rates were lower in the ILPS-treated-group. In addition, there were significantly fewer babies with an excessive ponderal index or neonatal hypoglycemic episodes in the ILPS group than in the NPH group. Conclusions Association of ILPS with rapid-acting analogs in pregnancy is safe in terms of maternal and fetal outcomes.
The journal of maternal-fetal & neonatal medicine: the official journal of the European Association of Perinatal Medicine, the Federation of Asia and Oceania Perinatal Societies, the International Society of Perinatal Obstetricians 03/2015; DOI:10.3109/14767058.2015.1033619 · 1.37 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: With real-time continuous glucose monitoring (RTCGM), the patient has a continuous picture of glucose concentrations - dynamic information with important predictive value. It is in fact a typical 'patient-oriented' tool that permits a more aggressive therapeutic approach and makes it easier to adapt therapy immediately. Currently approved RTCGM devices consist of three main components: A glucose sensor, a transmitter, and a receiver monitor. The system is defined as minimally invasive because it requires insertion into the subcutaneous cellular tissue of a fine needle to measure the glucose in the interstitial fluid directly or via an external sensor. A correct selection of patients is one of the most important factors in determining a successful application of this innovative technology. For this purpose, there are already consensus statements in a number of countries with regard to the indications of these systems. The criteria for identification of candidates are usually divided into three main categories: pediatric/adolescent, adult patients (both in the general population of type 1 diabetes and in specific settings), and pregnant women or women in the preconception period. RTCGM opens up extremely interesting prospects that are destined to radically change the concept of clinical self-management of diabetes. Considering the technical progress towards overcoming the current limitations (essentially: not optimal accuracy, invasiveness, and costs) which currently prevent its widespread use, we can expect wide diffusion of this form of monitoring in the coming years.
Frontiers in Diabetes 01/2015; 24:110-127. DOI:10.1159/000363482
[Show abstract][Hide abstract] ABSTRACT: Background: Continuous subcutaneous insulin infusion (CSII) is increasing worldwide, mostly because of improved technology. The aim of this study was to evaluate the current status of CSII in Italy.
Materials and Methods: Physicians from 272 diabetes centers received a questionnaire investigating clinical features, pump technology, and management of patients on CSII.
Results: Two hundred seventeen centers (79.8%) joined the study and, by the end of April 2013, gave information about 10,152 patients treated with CSII: 98.2% with type 1 diabetes mellitus, 81.4% adults, 57% female, and 61% with a conventional pump versus 39% with a sensor-augmented pump. CSII advanced functions were used by 68% of patients, and glucose sensors were used 12 days per month on average. Fifty-eight percent of diabetes centers had more than 20 patients on CSII, but there were differences among centers and among regions. The main indication for CSII was poor glucose control. Dropout was mainly due to pump wearability or nonoptimal glycemic control. Twenty-four hour assistance was guaranteed in 81% of centers. A full diabetes team (physician+nurse+dietician+psychologist) was available in 23% of adult-care diabetes centers and in 53% of pediatric diabetes units.
Conclusions: CSII keeps increasing in Italy. More work is needed to ensure uniform treatment strategies throughout the country and to improve pump use.
[Show abstract][Hide abstract] ABSTRACT: Increased blood glucose (BG) levels is common in critically ill patients both with and without a previous history of diabetes. Morbidity and mortality related to hyperglycemia is an important issue when managing BG in critically ill patients. Thus far, however, the characteristics of patients who may benefit from intensive insulin therapy remain to be clearly defined, as does the effect of different BG algorithms, the method of measuring BG, and the influence of different nutritional strategies. Intensive care units (ICUs) around the world have adopted insulin infusion protocols to achieve stringent BG targets in critically ill patients. Different insulin algorithms have been developed with variable effectiveness. Most of those algorithms require considerable training and experience. Computer decision support systems might improve the outcomes of patients in the ICU. Intensive insulin infusion managed using software-guided programs achieves tighter glycemic control and equal or fewer hypoglycemic episodes than paper protocols. The measurement of BG concentrations in the ICU is performed intermittently in the majority of ICUs, using either arterial blood gas analyzers, glucose meters, or central laboratory. Blood gas analyzers are considered the best compromise between accuracy and practicality. Real-time continuous glucose monitoring (RTCGM) offers interesting possibilities: automatic measurement of large numbers of BG values and alarms that warn whenever ‘outlier' values are reached. Data published to date from some randomized studies in ICUs have indicated a lower incidence of hypoglycemia when RTCGM is used in adults under mechanical ventilation and after cardiac surgery; however, until further studies provide sufficient evidence for its accuracy and safety, the use of RTCGM alone in not recommended.
[Show abstract][Hide abstract] ABSTRACT: The best way to treat pregnant patients who have type 1 diabetes is still unclear. For this reason, the present study compared metabolic control and maternal-fetal outcomes in patients treated with continuous subcutaneous infusions of rapid-acting insulin analogues (CSII) or with insulin glargine and multiple daily injections of rapid-acting insulin analogues (glargine-MDI).
This retrospective multicentre study involved 144 women with type 1 diabetes, 100 of whom were using CSII and 44 glargine-MDI. Outcomes analyzed were metabolic control, diabetes complications, pregnancy outcome, perinatal morbidity and mortality, and fetal malformations.
The two groups were comparable for age, prepregnancy BMI, primiparous rate and diabetes complications, although patients using CSII had longer duration of diabetes (P=0.03) and higher White classifications (P=0.04). In both groups, metabolic control improved during pregnancy, but good control was reached earlier among patients using CSII. At parturition, patients using CSII had lower HbA(1c) (6.2±0.7% vs 6.5±0.8%; P=0.02) and required less insulin (P<0.01). Weight gain was similar in both groups, and maternal-fetal outcomes did not differ.
In pregnant patients with type 1 diabetes, MDI and CSII are equivalent in terms of metabolic control and fetal-maternal outcomes, although patients using CSII achieved good control earlier and with less insulin.
[Show abstract][Hide abstract] ABSTRACT: This multicentre study analyzed the maternal and fetal outcomes of women who had one elevated 3-h oral glucose tolerance test (isolated gestational hyperglycaemia [IGH]).
From 1999 to 2003, data were collected for 606 IGH women from 31 Italian obstetric or diabetic centres, including time and mode of delivery, gestational hypertension, preeclampsia, eclampsia, congenital malformations, and neonatal mortality and morbidity, to compare them with the general pregnant Italian population. A prognostic model for the outcome of pregnancy was constructed, and the concurrence of certain specified conditions was considered a positive outcome, whereas pregnancies that failed to meet one or more of the stated conditions were classified as "complicated".
Macrosomia was significantly more frequent in women with IGH than in the normal pregnant population (10.7 vs 7.4%, respectively; P=0.003). Stillbirth and neonatal mortality rates did not differ from those in normal pregnancies, while a slight rise in the frequency of major malformations was not statistically significant (1.48 vs 0.89%, respectively; P<0.11). Multivariate logistic analyses confirmed that the prepregnancy body mass index (BMI) was an independent predictor of a complicated pregnancy. As for fetal growth, multivariate logistic analyses according to BMI showed that being overweight or obese were strong predictors of macrosomia.
These findings in a large cohort of Italian women with IGH confirm the detrimental effect of even minimally altered glucose tolerance on fetal outcome. Also, prepregnancy obesity plays an important role in raising the risk of adverse perinatal outcomes in such patients.
[Show abstract][Hide abstract] ABSTRACT: AimTo verify whether, with thorough practical and theoretical training, well-controlled, non-complicated diabetic patients can safely go diving underwater with no additional medical or metabolic risks.
[Show abstract][Hide abstract] ABSTRACT: To compare the effect of continuous subcutaneous insulin infusion (CSII) and multiple daily insulin injections (MDI) on albumin excretion rate (AER) in Type 1 diabetic patients.
In a 3-year multicentre retrospective observational study, 110 Type 1 diabetic patients treated with CSII were compared with 110 patients treated with MDI matched at baseline for age, sex, diabetes duration and HbA(1c). At entry, 90 patients in each group had normal AER and 20 persistent microalbuminuria. AER, estimated glomerular filtration rate (eGFR), HbA(1c,) lipids and blood pressure were assessed.
HbA(1c) was lower in the CSII than in the MDI group (8.1 +/- 0.9 vs. 8.4 +/- 1.3%; P < 0.005 after 3 years). Blood pressure and eGFR were similar during the study. AER [median (95% confidence interval)], similar at baseline [6.0 microg/min (9, 21) in the CSII group vs. 4.4 (8, 16) in the MDI group, NS] was significantly lower in the patients treated with CSII both at year 2 and at year 3 of follow-up [4.7 microg/min (6, 12) vs. 6.4 (13, 29), P < 0.002]. This difference was observed even when normo- and microalbuminuric patients were analysed separately. Nine patients progressed to microalbuminuria in the MDI group and only one in the CSII group. Nine patients regressed to normoalbuminuria in the CSII group, whereas only two regressed to normoalbuminuria in the MDI group.
Despite a small benefit in terms of improved glycaemic control, CSII therapy may be useful in decreasing the progressive increase in AER in Type 1 diabetic patients.
Diabetic Medicine 06/2009; 26(6):602-8. DOI:10.1111/j.1464-5491.2009.02736.x · 3.12 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: This prospective study evaluated the impact of gestational diabetes on maternal and fetal outcome in a large cohort of women with gestational diabetes mellitus (GDM) followed up using standardized clinical criteria.
Between 1999 and 2003, we collected 3465 GDM women from 31 Italian regional obstetric or diabetes centers, recording the time and mode of delivery, gestational hypertension, pre-eclampsia, eclampsia, congenital malformations, and neonatal mortality, comparing findings with the Italian general pregnant population.
The rate of cesarean sections was 34.9% and macrosomia 8.7% (33.2 and 7.4%, respectively, in the general population, p=ns). The stillbirth and neonatal mortality rates were no different in GDM patients and normal pregnancies (0.34% vs. 0.30%, p=0.176 and 0.29% vs. 0.32%, p=0.748), but the former had twice as many newborn with congenital malformations (2.05% vs. 0.89%, p<0.01; CI 1.64-2.62). A prognostic model for the outcome of pregnancy was built and the concurrent occurrence of several conditions was deemed as a positive outcome. Pregnancies which did not meet one or more of the above criteria were classified as "complicated". On multivariate logistic analysis, only the week of gestation when GDM was diagnosed and prepregnancy BMI were independent predictors of a complicated pregnancy.
When correctly diagnosed and treated during pregnancy, women with GDM have a pregnancy outcome similar to the general pregnant population, except for a greater likelihood of congenital malformations in the newborn, probably due to unrecognized prior diabetes. Prepregnancy obesity plays an important part in raising the risk of adverse perinatal outcomes in GDM patients.
European journal of obstetrics, gynecology, and reproductive biology 05/2009; 145(2):149-53. DOI:10.1016/j.ejogrb.2009.04.023 · 1.70 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Aim:
To verify whether, with thorough practical and theoretical training, well-controlled, non-complicated diabetic patients can safely go diving underwater with no additional medical or metabolic risks.
Twelve diabetic patients participated in the study after undergoing training focused on their diabetic status. Two dives per day were scheduled during two five-day stays on the island of Ventotene (Italy). Capillary blood glucose (BG) was checked at 60, 30 and 10 minutes before diving, and corrective measures adopted if necessary, based on BG absolute levels and dynamics. A device for continuous subcutaneous glucose monitoring (CGM), expressly modified for the purpose, was worn during dives.
Data were gathered from 90 dives; mean BG at 60, 30 and 10 minutes before diving was 205.8+/-69.6 mg/dL, 200.0+/-66.4 mg/dL and 200.5+/-61.0mg/dL, respectively. In 56 of the 90 dives, supplementary carbohydrates or insulin were necessary, but only one dive was interrupted on account of hypoglycaemic symptoms. Mean post-dive BG was 158.9+/-80.8 mg/dL. CGM recordings showed that glucose levels gradually decreased during the dives (nadir: -19.9%).
Experienced, well-controlled, complication-free young diabetic patients can safely go scuba diving, provided that they apply a rigorous protocol based on serial pre-dive BG measurements. The specific variables of underwater diving do not appear to involve significant additional risks of hypoglycaemia.
[Show abstract][Hide abstract] ABSTRACT: Insulin glargine (IG), with its non-peaking action profile, might be useful in diabetic pregnancy. However, data on its safety are limited and its use during pregnancy is not recommended. This study focused on the effects of IG on perinatal outcome, particularly to estimate the rate of congenital anomalies and birthweight.
This retrospective study included women with pre-gestational diabetes who used IG before (at least 1 month) and during pregnancy. For all women we recorded data regarding maternal glycaemic control and pregnancy outcome. We also compared women treated with IG throughout pregnancy and women who stopped taking IG at an earlier stage.
From 27 centres, 107 Type 1 diabetic pregnancies were identified. IG was started 10.3 +/- 6.9 months before conception and in 57.4% of cases was stopped during the first trimester; 42.6% of women continued using it until the end of pregnancy. There were six abortions (four spontaneous and two induced) and five newborns (4.9%) with congenital anomalies. Glycaemic control, birthweight and the prevalence of macrosomia and neonatal morbidity were similar in women who used IG for the full term compared with those who stopped IG earlier during pregnancy.
This study, although limited, suggests that IG is safe and effective; the rate of congenital malformations was within the range expected for diabetic pregnancies treated with more traditional forms of insulin. IG used throughout pregnancy did not seem to influence birthweight or increase adverse outcomes.
Diabetic Medicine 09/2008; 25(8):993-6. DOI:10.1111/j.1464-5491.2008.02485.x · 3.12 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: To determine pregnancy outcome in women with type 1 and type 2 diabetes.
A prospective study was conducted in 33 centers in Italy between 1999 and 2003, mainly recording preterm delivery, stillbirths, neonatal mortality, congenital malformations and birthweight. Of the 668 women examined, 504 had type 1 diabetes and 164 had type 2. Pre-pregnancy counseling had been provided to 43.9% of the women who had type 1 diabetes and 29.1% of the women who had type 2 diabetes and correlated with a better HbA1c value throughout pregnancy. The preterm delivery rate was significantly higher in type 1 and 2 diabetics than in normal pregnant women and was related to HbA1c values higher than 8%, gestational hypertension, pre-eclampsia and the presence of retinopathy before pregnancy. The stillbirth and neonatal mortality rates were also higher in diabetic pregnant women (1.26% and 0.63%, respectively) than in Italian pregnancies in general (0.30% and 0.32%), and the same was true for major congenital malformations (4.9% for diabetic pregnancies, 0.86% for normal Italian pregnancies).
In our population, pregnancy in diabetic women was still associated with a high rate of stillbirths, neonatal mortality and congenital malformations. Unplanned pregnancies and non-optimal glycemia control may help explain the high rates of maternal and neonatal complications.
[Show abstract][Hide abstract] ABSTRACT: Some studies have shown that fetal outcome observed in patients using insulin lispro is much the same as in pregnant women using regular insulin. This study aims to analyze the Italian data emerging from a multinational, multicenter, retrospective study on mothers with type 1 diabetes mellitus before pregnancy, comparing those treated with insulin lispro for at least 3 months before and 3 months after conception with those treated with regular insulin. The data collected on pregnant women with diabetes attending 15 Italian centers from 1998 to 2001 included: HbA1c at conception and during the first and third trimesters, frequency of severe hypoglycemic episodes, spontaneous abortions, mode and time of delivery, fetal malformations and mortality. Seventy-two diabetic pregnancies treated with lispro and 298 treated with regular insulin were analyzed, revealing a trend towards fewer hypoglycemic episodes in the former, who also had a significantly greater reduction in HbA1c during the first trimester. The rate of congenital malformations was similar in the offspring of the two groups of women treated with insulin lispro or regular insulin. These findings suggest that insulin lispro could be useful for the treatment of hyperglycemia in type 1 diabetic pregnant women.
[Show abstract][Hide abstract] ABSTRACT: To assess whether HbA1c and plasma glucose predicts abnormal fetal growth, 758 pregnant women attending 5 Diabetic Centers were screened for gestational diabetes mellitus (GDM). On glucose challenge (GCT) at 24-27 weeks of gestation (g.w.), negative cases formed the normal control group (N1). Positive cases took an oral glucose tolerance test (OGTT): those found negative were classed as false positives screening test (N2); if they had an OGTT result at least as high as their normal glucose levels, they were classed as having one abnormal glucose value (OAV) at OGTT; two values as GDM. HbA1c was assayed on the day of GCT. We considered fetal macrosomia, large for gestational age (LGA), ponderal index and mean growth percentile. Mean age, pre-pregnancy BMI, fasting plasma glucose (FPG) and HbA1c were progressively higher from N1 to GDM patients. The newborn of N2 mothers were heavier than those with N1 or GDM. The mean growth percentile was significantly higher in N2 than in N1. More LGA babies were born to OAV than to N1 or N2 women. Macrosomia and ponderal index did not differ significantly in the four groups. At logistic regression only plasma glucose at GCT could predict LGA babies and a ponderal index above 2.85. At risk analysis, GDM and OAV significantly predicted LGA babies, and GDM a ponderal index >2.85. In conclusion, FPG at GCT could predict fetal overgrowth and plasma glucose >85mg/dl doubles the risk of LGA infants. HbA1c at 24-27g.w. does not predict fetal overgrowth. Mild alterations in glucose tolerance correlate with fetal overgrowth and needs monitoring and treatment.
Diabetes Research and Clinical Practice 09/2007; 77(3):465-70. DOI:10.1016/j.diabres.2007.01.022 · 2.54 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: The reference intervals for hemoglobin A1c (Hb A1c) in pregnant women without diabetes are not well defined, and few examples of reference intervals established by networks of different laboratories are available.
Five Italian Diabetic Care Units were involved in the study. Data were collected from 445 pregnant women without diabetes, selected on the basis of glucose challenge test results, and from 384 nonpregnant control women. The Hb A1c measurements were performed with HPLC systems aligned to the Diabetes Control and Complications Trial. Plasma glucose measurements were also performed locally. Both Hb A1c and glucose measurements were harmonized by running appropriate external quality assessment schemes. The reference intervals were calculated in terms of nonparametric 2.5th to 97.5th percentiles with 0.90 confidence intervals.
The Hb A1c measurements were reproducible (CV = 2.0%) and accurate [mean (SE) difference from the target values, -0.10 (0.06)%]. Glucose measurements were also reproducible (mean CV = 3.2%) and accurate [difference from the target values, -0.01 (0.04) mmol/L]. To calculate common reference intervals, we merged the data collected in the different centers. The Hb A1c reference intervals were 4.0%-5.5% for pregnant nondiabetic women and 4.8%-6.2% for nonpregnant controls.
Healthy pregnant women have lower Hb A1c concentrations than nonpregnant women. The reference intervals for Hb A1c in pregnant women should therefore be lower than those currently in use.
[Show abstract][Hide abstract] ABSTRACT: Most studies relating minor gestational metabolic alterations to macrosomia refer to glucose intolerance classified on the basis of the National Diabetes Data Group or previous World Health Organization diagnostic thresholds. Our aim was to evaluate the consequences of very mild forms of gestational glucose intolerance, defined by an elevated 50-g glucose challenge test followed by a normal oral glucose tolerance test, using the more restrictive Carpenter and Coustan's criteria (Borderline Gestational Glucose Intolerance, BGGI).
Three hundred BGGI women were randomly assigned to: Group A (standard management), Group B (dietary treatment and regular monitoring). A control group (C) was also considered. Newborns were classified as macrosomic, large (LGA), or small for gestational age (SGA).
The three groups were similar in age, body mass index and parity. Therapy in Group B significantly improved fasting (from 4.68 +/- 0.45 to 4.28 +/- 0.45 mmol/l) and 2-h postprandial glycaemia (from 6.01 +/- 0.57 to 5.13 +/- 0.68 mmol/l). Fasting glycaemia at delivery was significantly lower in B (4.20 +/- 0.38 mmol/l) than in A (4.84 +/- 0.45 mmol/l), and was also lower than in C (4.31 +/- 0.39 mmol/l). Significantly fewer LGA babies were born to Group B (6.0%) than Group A (14.0%) and Group C (9.1%). No difference was found in the SGA rate. The neonatal Ponderal Index was higher (P = 0.030) in group A (2.73 +/- 0.35) than in C (2.64 +/- 0.30) and B (2.64 +/- 0.24).
Even very mild alterations in glucose tolerance can result in excessive or disharmonious fetal growth, which may be prevented by simple, non-invasive therapeutic measures.
Diabetic Medicine 12/2005; 22(11):1536-41. DOI:10.1111/j.1464-5491.2005.01690.x · 3.12 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: In order to prevent abnormalities of fetal growth still characterizing pregnancies complicated by Gestational Diabetes (GDM), in the present study we evaluated a therapeutic strategy for GDM based on ultrasound (US) measurement of fetal insulin-sensitive tissues.
All GDM women diagnosed before 28th week immediately started diet and self-monitoring of blood glucose; after 2 weeks they were randomized to conventional (C) or modified (M) management. In C the glycemic target (GT) was fixed at 90 fasting/120 post-prandial mg/dl; in M GT varied, according to US measurement of the Abdominal Circumference (AC) centile performed every 2 weeks: 80/100 if AC > or =75th, 100/140 if AC<75th. Therapy was tailored to mean fasting (FG) and postprandial glycemia (PPG).
Globally, 229 women completed the study, 78 in C, 151 in M. Use of insulin was 16.7% in C, 30.4% in M (total groups), significantly more frequent in M than in C (59.7% vs 15.4%) when considering only women with AC > or =75th c. Mean metabolic data were similar in the 2 groups, but in M a tightly-optimized subgroup, resulting from the lowering of GT due to AC > or =75th, coexisted with a less-controlled one, whose higher GT was justified by AC<75th. Pregnancy outcome was better in M, with lower (p<0.05*) rate of LGA* (7.9% vs 17.9%), SGA (6.0% vs 9.0%) and Macrosomia* (3.3% vs 11.5%).
Our data show the value of a flexible US-based approach to the treatment of GDM. This model does not necessarily involve a generalized aggressive treatment, allowing to concentrate therapeutical efforts on a small subgroup of women showing indirect US evidence of fetal hyperinsulinization. Such a selective approach allowed to obtain a near-normalization of fetal growth, with clear advantages on global pregnancy outcome.